AMD EPYC 4004 CPU Review Benchmarks

Benchmarks preparing for a future article on Phoronix.com on launch day by Michael Larabel.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2405204-PTS-EPYC400480
Jump To Table - Results

View

Do Not Show Noisy Results
Do Not Show Results With Incomplete Data
Do Not Show Results With Little Change/Spread
List Notable Results
Show Result Confidence Charts

Limit displaying results to tests within:

AV1 2 Tests
C++ Boost Tests 4 Tests
Chess Test Suite 2 Tests
Timed Code Compilation 10 Tests
C/C++ Compiler Tests 18 Tests
Compression Tests 3 Tests
CPU Massive 34 Tests
Creator Workloads 26 Tests
Cryptography 4 Tests
Database Test Suite 7 Tests
Encoding 9 Tests
Fortran Tests 6 Tests
Game Development 6 Tests
HPC - High Performance Computing 20 Tests
Imaging 3 Tests
Java 3 Tests
Common Kernel Benchmarks 4 Tests
Machine Learning 6 Tests
Molecular Dynamics 7 Tests
MPI Benchmarks 7 Tests
Multi-Core 50 Tests
NVIDIA GPU Compute 4 Tests
Intel oneAPI 7 Tests
OpenMPI Tests 12 Tests
Programmer / Developer System Benchmarks 13 Tests
Python 2 Tests
Raytracing 2 Tests
Renderers 6 Tests
Scientific Computing 10 Tests
Software Defined Radio 2 Tests
Server 13 Tests
Server CPU Tests 25 Tests
Single-Threaded 3 Tests
Texture Compression 2 Tests
Video Encoding 9 Tests
Common Workstation Benchmarks 3 Tests

Statistics

Show Overall Harmonic Mean(s)
Show Overall Geometric Mean
Show Geometric Means Per-Suite/Category
Show Wins / Losses Counts (Pie Chart)
Normalize Results
Remove Outliers Before Calculating Averages

Graph Settings

Force Line Graphs Where Applicable
Convert To Scalar Where Applicable
Disable Color Branding
Prefer Vertical Bar Graphs
No Box Plots
On Line Graphs With Missing Data, Connect The Line Gaps

Additional Graphs

Show Perf Per Core/Thread Calculation Graphs Where Applicable
Show Perf Per Clock Calculation Graphs Where Applicable

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs
Condense Test Profiles With Multiple Version Results Into Single Result Graphs

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Hide
Result
Result
Identifier
View Logs
Performance Per
Dollar
Date
Run
  Test
  Duration
Xeon E-2488
April 01
  1 Day, 22 Hours, 13 Minutes
EPYC 4244P
May 12
  1 Day, 16 Hours, 20 Minutes
EPYC 4344P
May 08
  1 Day, 11 Hours, 13 Minutes
EPYC 4364P
April 28
  1 Day, 17 Hours, 57 Minutes
EPYC 4464P
May 17
  1 Day, 11 Hours, 26 Minutes
EPYC 4484PX
May 10
  1 Day, 14 Hours, 16 Minutes
EPYC 4564P
April 27
  1 Day, 8 Hours, 35 Minutes
EPYC 4584PX
May 15
  1 Day, 9 Hours, 50 Minutes
Invert Hiding All Results Option
  1 Day, 13 Hours, 59 Minutes

Only show results where is faster than
Only show results matching title/arguments (delimit multiple options with a comma):
Do not show results matching title/arguments (delimit multiple options with a comma):


AMD EPYC 4004 CPU Review BenchmarksProcessorMotherboardChipsetMemoryDiskGraphicsNetworkAudioOSKernelDesktopDisplay ServerCompilerFile-SystemScreen ResolutionXeon E-2488EPYC 4564PEPYC 4364PEPYC 4344PEPYC 4484PXEPYC 4244PEPYC 4584PXEPYC 4464PIntel Xeon E E-2488 @ 3.20GHz (8 Cores / 16 Threads)Supermicro Super Server X13SCL-F v0123456789 (1.1 BIOS)Intel Raptor Lake-S PCH2 x 32GB DRAM-4800MT/s Micron MTC20C2085S1EC48BA13201GB Micron_7450_MTFDKCC3T2TFS + 960GB Micron_7450_MTFDKBA960TFRASPEED2 x Intel I210Ubuntu 24.046.8.0-11-generic (x86_64)GNOME Shell 45.3X Server 1.21.1.11GCC 13.2.0ext41920x1080AMD EPYC 4564P 16-Core @ 5.88GHz (16 Cores / 32 Threads)Supermicro AS-3015A-I H13SAE-MF v1.00 (1.1 BIOS)AMD Device 14d82 x 32GB DRAM-5200MT/sAMD Rembrandt Radeon HD AudioAMD EPYC 4364P 8-Core @ 5.57GHz (8 Cores / 16 Threads)AMD EPYC 4344P 8-Core @ 5.39GHz (8 Cores / 16 Threads)AMD EPYC 4484PX 12-Core @ 5.66GHz (12 Cores / 24 Threads)AMD EPYC 4244P 6-Core @ 5.17GHz (6 Cores / 12 Threads)AMD EPYC 4584PX 16-Core @ 5.76GHz (16 Cores / 32 Threads)AMD EPYC 4464P 12-Core @ 5.48GHz (12 Cores / 24 Threads)OpenBenchmarking.orgKernel Details- Transparent Huge Pages: madviseCompiler Details- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-fxIygj/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-fxIygj/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v Disk Details- Xeon E-2488: NONE / errors=remount-ro,relatime,rw,stripe=64 / Block Size: 4096Processor Details- Xeon E-2488: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0x122 - Thermald 2.5.6- EPYC 4564P: Scaling Governor: amd-pstate-epp performance (EPP: performance) - CPU Microcode: 0xa601203- EPYC 4364P: Scaling Governor: amd-pstate-epp performance (EPP: performance) - CPU Microcode: 0xa601203- EPYC 4344P: Scaling Governor: amd-pstate-epp performance (EPP: performance) - CPU Microcode: 0xa601203- EPYC 4484PX: Scaling Governor: amd-pstate-epp performance (EPP: performance) - CPU Microcode: 0xa601203- EPYC 4244P: Scaling Governor: amd-pstate-epp performance (EPP: performance) - CPU Microcode: 0xa601203- EPYC 4584PX: Scaling Governor: amd-pstate-epp performance (EPP: performance) - CPU Microcode: 0xa601203- EPYC 4464P: Scaling Governor: amd-pstate-epp performance (EPP: performance) - CPU Microcode: 0xa601203Java Details- OpenJDK Runtime Environment (build 21.0.2+13-Ubuntu-2)Python Details- Python 3.12.2Security Details- Xeon E-2488: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected - EPYC 4564P: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Vulnerable: Safe RET no microcode + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 4364P: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Vulnerable: Safe RET no microcode + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 4344P: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Vulnerable: Safe RET no microcode + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 4484PX: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Vulnerable: Safe RET no microcode + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 4244P: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Vulnerable: Safe RET no microcode + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 4584PX: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Vulnerable: Safe RET no microcode + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 4464P: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Vulnerable: Safe RET no microcode + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

Xeon E-2488EPYC 4564PEPYC 4364PEPYC 4344PEPYC 4484PXEPYC 4244PEPYC 4584PXEPYC 4464PResult OverviewPhoronix Test Suite100%176%252%328%404%NAMDminiBUDEJohn The RipperOpenSSLPrimesieveOSPRay StudioStockfishOpenVINOACES DGEMMPostgreSQLCpuminer-OptASTC Encoderm-queensAircrack-ngStress-NGJava JMHBlenderCoremarkQuantLibChaos Group V-RAYHelsingEmbreeLuxCoreRenderIntel Open Image DenoiseOpenVKLASKAPOSPRayMemcachedBRL-CADTimed Node.js CompilationTimed Linux Kernel CompilationParallel BZIP2 Compression7-Zip CompressionTimed LLVM CompilationTimed Godot Game Engine CompilationSVT-HEVCnginxAppleseedGROMACSTimed FFmpeg CompilationRocksDBuvg266oneDNNSPECFEM3DTimed Mesa CompilationSpeedbTimed Gem5 CompilationGraphicsMagickKvazaarHeFFTe - Highly Efficient FFT for ExascaleScyllaDBSVT-VP9LAMMPS Molecular Dynamics SimulatorNAS Parallel BenchmarksJPEG-XL libjxlCockroachDBCloverLeafY-Cruncherlibavif avifencPennantsrsRAN Projectx265DaCapo BenchmarkTensorFlowSVT-AV1Zstd CompressionTimed PHP CompilationOpenFOAMVVenCGPAWMariaDBOpenRadiossXcompact3d Incompact3dPyBenchONNX RuntimeTimed Wasmer CompilationDuckDBGoogle DracoClickHouseZ3 Theorem ProverLiquid-DSPApache IoTDBPyTorchRenaissanceTimed CPython CompilationNumpy BenchmarkGNU Octave BenchmarkMariaDB mariadb-slapFFmpeg

Xeon E-2488EPYC 4564PEPYC 4364PEPYC 4344PEPYC 4484PXEPYC 4244PEPYC 4584PXEPYC 4464PPer Watt Result OverviewPhoronix Test Suite100%163%225%288%350%NAMDminiBUDEJohn The Ripper7-Zip CompressionHeFFTe - Highly Efficient FFT for ExascaleACES DGEMMASKAPOpenSSLJava JMHEmbreeIntel Open Image DenoiseAircrack-ngChaos Group V-RAYNumpy BenchmarkCoremarkOpenVKLCpuminer-OptStockfishClickHouseOSPRayStress-NGGROMACSMemcachednginxASTC EncodersrsRAN ProjectKvazaaruvg266LAMMPS Molecular Dynamics SimulatorSVT-HEVCx265SVT-VP9TensorFlowNAS Parallel BenchmarksPyTorchLuxCoreRenderLiquid-DSPSVT-AV1JPEG-XL libjxlBRL-CADQuantLibRocksDBScyllaDBMariaDB mariadb-slapSpeedbGraphicsMagickFFmpegCockroachDBVVenCMariaDBP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.M

AMD EPYC 4004 CPU Review Benchmarkslammps: 20k Atomstensorflow: CPU - 512 - ResNet-50openradioss: Chrysler Neon 1Mbuild-linux-kernel: allmodconfigopenradioss: INIVOL and Fluid Structure Interaction Drop Containeropenfoam: drivaerFastback, Medium Mesh Size - Execution Timeopenfoam: drivaerFastback, Medium Mesh Size - Mesh Timetensorflow: CPU - 256 - ResNet-50blender: Barbershop - CPU-Onlybuild-nodejs: Time To Compileopenvkl: vklBenchmarkCPU ISPCtensorflow: CPU - 512 - GoogLeNetbuild-llvm: Unix Makefilesopenradioss: Bird Strike on Windshieldbuild-llvm: Ninjabuild-gem5: Time To Compileopenradioss: Rubber O-Ring Seal Installationopenradioss: Bumper Beammariadb: oltp_update_non_index - 64helsing: 14 digitapache-iotdb: 800 - 100 - 800 - 400apache-iotdb: 800 - 100 - 800 - 400gpaw: Carbon Nanotubeopenssl: RSA4096openssl: RSA4096apache-iotdb: 800 - 100 - 800 - 100apache-iotdb: 800 - 100 - 800 - 100openradioss: Cell Phone Drop Testtensorflow: CPU - 256 - GoogLeNetblender: Pabellon Barcelona - CPU-Onlybuild-godot: Time To Compilemariadb: oltp_read_write - 64mariadb: oltp_write_only - 64mariadb: oltp_update_index - 64stockfish: Chess Benchmarkclickhouse: 100M Rows Hits Dataset, Third Runclickhouse: 100M Rows Hits Dataset, Second Runclickhouse: 100M Rows Hits Dataset, First Run / Cold Cacheblender: Classroom - CPU-Onlytensorflow: CPU - 64 - ResNet-50tensorflow: CPU - 512 - AlexNetonnx: ResNet50 v1-12-int8 - CPU - Standardonnx: ResNet50 v1-12-int8 - CPU - Standardminibude: OpenMP - BM2minibude: OpenMP - BM2apache-iotdb: 800 - 100 - 500 - 400apache-iotdb: 800 - 100 - 500 - 400apache-iotdb: 500 - 100 - 800 - 400apache-iotdb: 500 - 100 - 800 - 400apache-iotdb: 800 - 100 - 500 - 100apache-iotdb: 800 - 100 - 500 - 100brl-cad: VGR Performance Metricapache-iotdb: 500 - 100 - 800 - 100apache-iotdb: 500 - 100 - 800 - 100svt-hevc: 1 - Bosphorus 4Kopenssl: ChaCha20openssl: AES-256-GCMopenssl: AES-128-GCMopenssl: ChaCha20-Poly1305openssl: SHA512openssl: SHA256mysqlslap: 128speedb: Read While Writingincompact3d: input.i3d 193 Cells Per Directionpgbench: 100 - 800 - Read Only - Average Latencypgbench: 100 - 800 - Read Onlyospray: particle_volume/scivis/real_timejava-jmh: Throughputcompress-zstd: 19 - Compression Speedpgbench: 100 - 800 - Read Write - Average Latencypgbench: 100 - 800 - Read Writeappleseed: Emilyscylladb: Writesonnx: fcn-resnet101-11 - CPU - Parallelonnx: fcn-resnet101-11 - CPU - Parallelastcenc: Very Thoroughastcenc: Exhaustiveapache-iotdb: 500 - 100 - 500 - 400apache-iotdb: 500 - 100 - 500 - 400ospray: particle_volume/pathtracer/real_timeprimesieve: 1e13apache-iotdb: 500 - 100 - 500 - 100apache-iotdb: 500 - 100 - 500 - 100pgbench: 100 - 500 - Read Only - Average Latencypgbench: 100 - 500 - Read Onlypgbench: 100 - 500 - Read Write - Average Latencypgbench: 100 - 500 - Read Writeonnx: yolov4 - CPU - Parallelonnx: yolov4 - CPU - Parallelnpb: SP.Cffmpeg: libx265 - Platformffmpeg: libx265 - Video On Demandffmpeg: libx265 - Uploadnpb: LU.Cduckdb: IMDBonnx: T5 Encoder - CPU - Standardonnx: T5 Encoder - CPU - Standardluxcorerender: Danish Mood - CPUblender: Junkshop - CPU-Onlyspecfem3d: Water-layered Halfspacespecfem3d: Layered Halfspaceospray: particle_volume/ao/real_timeonnx: GPT-2 - CPU - Standardonnx: GPT-2 - CPU - Standardoidn: RTLightmap.hdr.4096x4096 - CPU-Onlyblender: Fishy Cat - CPU-Onlytensorflow: CPU - 32 - ResNet-50namd: STMV with 1,066,628 Atomstensorflow: CPU - 256 - AlexNetavifenc: 0cockroach: KV, 50% Reads - 128cockroach: KV, 60% Reads - 128cockroach: KV, 95% Reads - 128onnx: bertsquad-12 - CPU - Parallelonnx: bertsquad-12 - CPU - Parallelvvenc: Bosphorus 4K - Fastgromacs: MPI CPU - water_GMX50_barenpb: BT.Cnumpy: ospray-studio: 3 - 1080p - 16 - Path Tracer - CPUmt-dgemm: Sustained Floating-Point Ratepytorch: CPU - 64 - ResNet-50compress-zstd: 12 - Compression Speedappleseed: Material Testerappleseed: Disney Materialaskap: tConvolve MPI - Griddingaskap: tConvolve MPI - Degriddingmysqlslap: 64blender: BMW27 - CPU-Onlynginx: 1000nginx: 500nginx: 200openfoam: drivaerFastback, Small Mesh Size - Execution Timeopenfoam: drivaerFastback, Small Mesh Size - Mesh Timebuild-linux-kernel: defconfigpennant: leblancbigospray-studio: 3 - 1080p - 32 - Path Tracer - CPUonednn: Recurrent Neural Network Training - CPUcloverleaf: clover_bmcompress-zstd: 8 - Compression Speedonnx: CaffeNet 12-int8 - CPU - Parallelonnx: CaffeNet 12-int8 - CPU - Parallelcompress-zstd: 19, Long Mode - Compression Speedonednn: Recurrent Neural Network Inference - CPUliquid-dsp: 1 - 256 - 512namd: ATPase with 327,506 Atomsospray-studio: 3 - 1080p - 1 - Path Tracer - CPUtensorflow: CPU - 64 - GoogLeNetspecfem3d: Mount St. Helensv-ray: CPUospray-studio: 2 - 1080p - 32 - Path Tracer - CPUonnx: ArcFace ResNet-100 - CPU - Parallelonnx: ArcFace ResNet-100 - CPU - Parallelospray-studio: 1 - 1080p - 32 - Path Tracer - CPUospray-studio: 1 - 1080p - 1 - Path Tracer - CPUospray-studio: 2 - 1080p - 1 - Path Tracer - CPUpennant: sedovbigospray: gravity_spheres_volume/dim_512/pathtracer/real_timebuild-python: Released Build, PGO + LTO Optimizedmemcached: 1:10memcached: 1:100uvg266: Bosphorus 4K - Mediumospray: gravity_spheres_volume/dim_512/ao/real_timeospray: gravity_spheres_volume/dim_512/scivis/real_timeonnx: Faster R-CNN R-50-FPN-int8 - CPU - Parallelonnx: Faster R-CNN R-50-FPN-int8 - CPU - Parallelastcenc: Thoroughluxcorerender: Orange Juice - CPUonnx: ResNet50 v1-12-int8 - CPU - Parallelonnx: ResNet50 v1-12-int8 - CPU - Parallelspecfem3d: Tomographic Modelospray-studio: 2 - 1080p - 16 - Path Tracer - CPUcompress-zstd: 8, Long Mode - Compression Speedopenvino: Face Detection FP16 - CPUopenvino: Face Detection FP16 - CPUluxcorerender: LuxCore Benchmark - CPUopenvino: Face Detection FP16-INT8 - CPUopenvino: Face Detection FP16-INT8 - CPUospray-studio: 1 - 1080p - 16 - Path Tracer - CPUonnx: GPT-2 - CPU - Parallelonnx: GPT-2 - CPU - Parallelgraphics-magick: Sharpenluxcorerender: DLSC - CPUopenvino: Machine Translation EN To DE FP16 - CPUopenvino: Machine Translation EN To DE FP16 - CPUonnx: T5 Encoder - CPU - Parallelonnx: T5 Encoder - CPU - Parallelopenvino: Person Detection FP16 - CPUopenvino: Person Detection FP16 - CPUopenvino: Person Detection FP32 - CPUopenvino: Person Detection FP32 - CPUgraphics-magick: Enhancedopenvino: Noise Suppression Poconet-Like FP16 - CPUopenvino: Noise Suppression Poconet-Like FP16 - CPUopenvino: Road Segmentation ADAS FP16-INT8 - CPUopenvino: Road Segmentation ADAS FP16-INT8 - CPUopenvino: Person Vehicle Bike Detection FP16 - CPUopenvino: Person Vehicle Bike Detection FP16 - CPUgraphics-magick: Noise-Gaussianopenvino: Road Segmentation ADAS FP16 - CPUopenvino: Road Segmentation ADAS FP16 - CPUopenvino: Person Re-Identification Retail FP16 - CPUopenvino: Person Re-Identification Retail FP16 - CPUopenvino: Handwritten English Recognition FP16-INT8 - CPUopenvino: Handwritten English Recognition FP16-INT8 - CPUopenvino: Vehicle Detection FP16-INT8 - CPUopenvino: Vehicle Detection FP16-INT8 - CPUopenvino: Handwritten English Recognition FP16 - CPUopenvino: Handwritten English Recognition FP16 - CPUopenvino: Face Detection Retail FP16-INT8 - CPUopenvino: Face Detection Retail FP16-INT8 - CPUspecfem3d: Homogeneous Halfspaceopenvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPUopenvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPUopenvino: Vehicle Detection FP16 - CPUopenvino: Vehicle Detection FP16 - CPUopenvino: Weld Porosity Detection FP16 - CPUopenvino: Weld Porosity Detection FP16 - CPUgraphics-magick: Swirlopenvino: Age Gender Recognition Retail 0013 FP16 - CPUopenvino: Age Gender Recognition Retail 0013 FP16 - CPUopenvino: Weld Porosity Detection FP16-INT8 - CPUopenvino: Weld Porosity Detection FP16-INT8 - CPUopenvino: Face Detection Retail FP16 - CPUopenvino: Face Detection Retail FP16 - CPUjohn-the-ripper: MD5speedb: Read Rand Write Randgraphics-magick: HWB Color Spacegraphics-magick: Resizingspeedb: Rand Readrocksdb: Read Rand Write Randrocksdb: Read While Writingonnx: super-resolution-10 - CPU - Parallelonnx: super-resolution-10 - CPU - Paralleljohn-the-ripper: HMAC-SHA512rocksdb: Rand Readavifenc: 2incompact3d: input.i3d 129 Cells Per Directionoidn: RT.hdr_alb_nrm.3840x2160 - CPU-Onlyoidn: RT.ldr_alb_nrm.3840x2160 - CPU-Onlyz3: 2.smt2build-php: Time To Compilepytorch: CPU - 1 - ResNet-50m-queens: Time To Solveembree: Pathtracer ISPC - Asian Dragon Objcoremark: CoreMark Size 666 - Iterations Per Secondrenaissance: Savina Reactors.IOkvazaar: Bosphorus 4K - Mediumsvt-hevc: 1 - Bosphorus 1080pvvenc: Bosphorus 4K - Fasterx265: Bosphorus 4Kffmpeg: libx265 - Livejpegxl: JPEG - 90jpegxl: PNG - 90embree: Pathtracer ISPC - Crownquantlib: Multi-Threadedbuild-ffmpeg: Time To Compiletensorflow: CPU - 32 - GoogLeNetnpb: SP.Bembree: Pathtracer ISPC - Asian Dragonbuild-wasmer: Time To Compilesvt-vp9: Visual Quality Optimized - Bosphorus 4Ksrsran: PUSCH Processor Benchmark, Throughput Threadtensorflow: CPU - 64 - AlexNetminibude: OpenMP - BM1minibude: OpenMP - BM1vvenc: Bosphorus 1080p - Fastaircrack-ng: john-the-ripper: WPA PSKjohn-the-ripper: bcryptjohn-the-ripper: Blowfishliquid-dsp: 16 - 256 - 512cpuminer-opt: LBC, LBRY Creditsstress-ng: Wide Vector Mathstress-ng: Vector Floating Pointcpuminer-opt: Magistress-ng: Floating Pointcpuminer-opt: Myriad-Groestlcpuminer-opt: Deepcoinstress-ng: Vector Shufflestress-ng: Fused Multiply-Addliquid-dsp: 16 - 256 - 32stress-ng: AVX-512 VNNIliquid-dsp: 1 - 256 - 32cpuminer-opt: Garlicoincpuminer-opt: scryptcpuminer-opt: Quad SHA-256, Pyritecpuminer-opt: Ringcoincpuminer-opt: Triple SHA-256, Onecoincpuminer-opt: Skeincoincpuminer-opt: Blake-2 Ssvt-av1: Preset 4 - Bosphorus 4Ksvt-av1: Preset 13 - Bosphorus 4Kjpegxl: JPEG - 100jpegxl: PNG - 100y-cruncher: 1Bbuild-mesa: Time To Compiledacapobench: H2 Database Engineuvg266: Bosphorus 4K - Very Fastpybench: Total For Average Test Timesastcenc: Fastuvg266: Bosphorus 4K - Super Fastnpb: EP.Ctensorflow: CPU - 32 - AlexNetsrsran: PUSCH Processor Benchmark, Throughput Totaldacapobench: Apache Tomcatonednn: Deconvolution Batch shapes_1d - CPUcompress-7zip: Decompression Ratingcompress-7zip: Compression Ratingkvazaar: Bosphorus 4K - Ultra Fastuvg266: Bosphorus 4K - Ultra Fastdacapobench: Apache Lucene Search Engineonednn: IP Shapes 1D - CPUkvazaar: Bosphorus 4K - Super Fastkvazaar: Bosphorus 4K - Very Fastaskap: tConvolve OpenMP - Degriddingaskap: tConvolve OpenMP - Griddingluxcorerender: Rainbow Colors and Prism - CPUsvt-vp9: VMAF Optimized - Bosphorus 4Kvvenc: Bosphorus 1080p - Fastery-cruncher: 500Mastcenc: Mediumuvg266: Bosphorus 1080p - Mediumcompress-pbzip2: FreeBSD-13.0-RELEASE-amd64-memstick.img Compressionsvt-av1: Preset 8 - Bosphorus 4Kdacapobench: PMD Source Code Analyzertensorflow: CPU - 1 - ResNet-50primesieve: 1e12x265: Bosphorus 1080psvt-hevc: 7 - Bosphorus 4Ksvt-vp9: PSNR/SSIM Optimized - Bosphorus 4Kavifenc: 6, Losslesssvt-av1: Preset 4 - Bosphorus 1080psvt-hevc: 10 - Bosphorus 4Konednn: IP Shapes 3D - CPUkvazaar: Bosphorus 1080p - Mediumdraco: Church Facadeuvg266: Bosphorus 1080p - Very Fastsvt-av1: Preset 12 - Bosphorus 4Konednn: Convolution Batch Shapes Auto - CPUuvg266: Bosphorus 1080p - Super Fastavifenc: 6srsran: PDSCH Processor Benchmark, Throughput Totalsvt-av1: Preset 8 - Bosphorus 1080poctave-benchmark: uvg266: Bosphorus 1080p - Ultra Fastkvazaar: Bosphorus 1080p - Very Fastdacapobench: Zxing 1D/2D Barcode Image Processingkvazaar: Bosphorus 1080p - Super Fastsvt-hevc: 7 - Bosphorus 1080phelsing: 12 digittensorflow: CPU - 1 - GoogLeNetlammps: Rhodopsin Proteinsvt-vp9: Visual Quality Optimized - Bosphorus 1080ponednn: Deconvolution Batch shapes_3d - CPUsrsran: PDSCH Processor Benchmark, Throughput Threadkvazaar: Bosphorus 1080p - Ultra Fastsvt-vp9: VMAF Optimized - Bosphorus 1080psvt-vp9: PSNR/SSIM Optimized - Bosphorus 1080psvt-hevc: 10 - Bosphorus 1080psvt-av1: Preset 12 - Bosphorus 1080pbuild-python: Defaultheffte: r2c - FFTW - double - 128heffte: r2c - FFTW - float - 128Xeon E-2488EPYC 4564PEPYC 4364PEPYC 4344PEPYC 4484PXEPYC 4244PEPYC 4584PXEPYC 4464P7.57922.98983.541508.042538.422260.9881225.9549123.071281.54915.93728876.13719.392259.88697.937468.378177.74153.90220068437.414524.1356494499228.899195618.83000.2137.925603527893.2676.84411.15364.42218641829603418708916898291249.90249.50234.39336.9423.10192.284.46170224.10612.550313.737290.1462655549382.486488636477.2461842388194127119.49625580982.2634651855800561932491906481272568023800263020442939716012336557263574254883773.72036131.8524319625.8864036765943472.26816.323.19834486323.270593104832960.8561.0428121.46390.8741275.1457760197157.969245.67480.56570283980.89955587311.79142412123.5938.0942111387.0558.7558.7528.8940491.5699.5755.90786169.6201.45173.17135.416735242144.5395371755.931617.89080126.6740.21165.0023.190.20565189.14133.56759866.663021.483987.586.687311.53724.9751.43428326.36675.48489294.20411032.25170.6173.408541174.9489676199.095146.331128123.5187132.2292582.2998692.95242.9447935.18608108.33832.86765954472931.9564.58485.61.79347557.4088.911548.71202550000.65977288478.1753.287162292200608240343.396723.0436812362441247455.896473.79539197.9213072373.873048853.566.622.734962.6385124.385841.009312.23073.584.66040214.54752.89378166342275504.11506.632.651.69384.0910.41415148.10864123.284362.13118.4533.746.19876161.313133.9929.83134.5429.714915.21523.0236.35219.869.94401.427039.48101.2118.22437.9645.70174.9215.95500.8653.07150.645.071559.0667.1149295080.418507.3116.04249.0230.05265.831301.007499.477.781022.109.76816.179197912120915315244578244261976533194276013.179275.8803407210004751226061.84319.75214900.420.4263.07654.44961.4972.85214.5499430017.4605833695.49.289.1611.76819.48178.1532.50631.81413.388337214.450.05178.7912750.9716.575738.13459.56153.6173.7812.452311.31015.46335441.7072956018787188411592700006037.26576480.1528846.06278.094740.284843.723294.499138.2929762395.577110500001125583.01632366671356.31119.86265271437.473884713933561604.348104.61413.48413.65033.89132.515212818.97415205.932120.171470.81160.521068.966876.85318615279499039.5924.6931132.9140828.6423.054282.182556.229.3579.8540.06714.92689.140036.596.61248636.140112213.7121.09584.3347.9883.198.96915.918101.837.5969550.47560789.31105.2806.3805396.336.2602908.5120.5336.462115.40103.041053139.72168.394.08051.337.236221.334.94792809.6197.51289.26295.69383.64548.04216.8526.478785.006213.69141.02891.84714.384322.672351.5135226.0110440.68481.15421.430603132.62387.011183.64354.764273.493110.05108.64233313200.768399.5673716025154.989373632.314289.7103.377407524467.71132.13165.35188.52520118234809919853844270858251.80248.89227.39136.1140.24425.472.38299422.41749.9851249.619235.1677658914321.297845431660.667760562739713892.08793153195.80129853526473215336847677251224402513921296221501104157992333989652273612511989173.88367840.72211082438.9819575134161573.04218.816.15749522152.301783151729567.0651.763492.68411.6391230.4169731166220.90782.47365.63698575190.36113841668.8505651786.840311.519913374.3569.8069.8134.2147556.6982.3776.16145162.3294.1872.7585.25437026083.9703858868.983997.51744133.1680.4166.7839.990.75646409.9672.10692195.295084.8121061.764.282715.55607.8182.54344844.13803.04202839.57775343.38220.9102.15238185.3179227477.196642.94118353.25136360.72154835.06162348.33183.2345130.76527157.47229.55297442471348.3772.88793.81.20504829.19410.7706.645189644002.788461265133.2133.126776518438793819226.828437.2714379231069108042.225658.97813193.4653761626.613844031.2815.927.466497.3690918.966452.721419.90387.722.62524380.81830.13223158017285797.0590.8013.524.65301.7926.46171168.15155122.627585.0058.43136.736.37806156.77389.9888.8490.4988.32949.901587.7416.73476.635.121556.0711925.15317.254.111938.3220.09795.054.821653.6921.89729.843.334734.2739.5610713750.2949485.3511.36701.8411.721362.002500.4134815.395.962672.512.373269.1241180002770195278338134971626291575842946317.50080133.32218133100013069306336.33116.58846850.860.8648.87740.11065.6927.91427.3873951674.5064613265.222.0223.1017.43534.65186.4248.58848.52531.365883365.125.301135.3018660.9032.220929.443100.00183.5322.5349.9531248.81023.19092259.953160664436954379529239000016120778238.3359037.14633.448199.65113378356.7416452.4756410032.257959666672571394.82530186673179.32308.21613603500.1397340346401427537.838165.28527.22327.74818.21117.464182642.60533392.237746.412123.37242.491611.533703.0063617037317088881.2354.5420801.1300062.5648.115172.445073.2316.98103.5250.9148.652154.720264.163.29987972.744105314.936.691118.0699.61111.015.89024.171158.943.8868588.305224166.81170.2146.40719181.833.42112852.6189.4135.780203.39186.56589248.39311.952.01657.3412.481364.902.461591049.1301.57408.42419.26560.81660.33614.13437.2403127.9318.21530.701003.591310.962585.752827.8789296.2109430.63925.38739.34333191.90662.522266.65636.832432.221162.03145.32201401375.915462.1464182563205.874194210.37697.3122.556295145895.4891.77315.13326.16216859326511917757923143858232.47236.47222.47259.4630.53240.233.09150323.38126.357658.931268.4269144164347.977133874068.1669561449226198105.58703052853.046836610881011394125887713313407467348542739663579440530717876358337580319497187.77034252.0513900184.6187138776376503.46518.024.49832657263.51109894217864.5611.160541.39790.8557243.6164655502170.390153.59870.70647812850.96152050212.35540473107.7939.2917910027.1165.2665.3531.9537149.88100.6277.16751139.4772.10139.48129.332051612133.3090441854.638288.06389123.9520.23127.0130.790.49345237.57108.65464052.767506.783155.879.703712.54835.6931.54622935.27808.62438305.33050045.11171.0142.936958156.0695324256.283635.691124101.7285990.1294947.75101214.88296.6009838.17300497.40542.93941842092491.1486.82520.31.47531677.58310.21337.47196116671.73764252692.3751.286977035230787225037.002527.0247715302133215661.969164.66757200.9782195688.492104490.558.613.910703.8213819.791050.531010.20894.423.09905322.61049.57647068837900527.9560.147.142.41285.9713.97374588.31107120.280302.7548.2682.807.06097141.62972.5255.1172.7054.98507.95991.3214.48275.724.78833.737617.79224.433.871029.8819.13417.644.53879.5020.37392.233.162495.5563.8652030420.2826230.858.44472.5711.05723.131390.418615.125.631417.332.211781.6222343332138776255255716486832131122251959410.918291.58551167703336940825451.94221.87464310.480.4848.67653.75567.4853.50115.6889499103.5029053042.012.4912.2813.14117.61179.9937.18036.02416.462543490.742.19593.7111152.3118.451136.90261.90183.6206.4026.294657.34518.77348527.8158553223235232432116700008389.15413926.8230936.50330.624251.395972.094377.748556.7229927440.697355666671371242.64536590001674.99162.88319031851.255108717983745135.742122.90816.51916.66125.45528.263197725.25527203.103926.651169.68172.891470.564823.910898894510357152.3031.8624091.9998638.8231.043426.032739.3411.9371.4845.14411.58979.604045.636.15446047.285111317.4712.50689.1257.2376.068.14719.891109.034.6784064.685083111.78125.2778.16800119.835.3387530.0148.3365.587142.97129.021019168.08178.203.82862.996.826203.163.834121061.5226.13242.58251.44361.05546.70616.57623.717977.51968.95130.49967.201319.932494.792781.2362262.8387330.50942.29743.11132790.13669.644238.13644.770438.615127.32124.70197477386.155465.5163954176203.432190537.17335.2122.336308624780.0490.19323.68330.82416606026013917308021844343233.30230.37213.47267.8630.64238.892.91846343.17925.885647.131260.6070107030350.007120389368.5869342377221767105.70702935712.946657277460711042558725312904662694047254230173564435758317402712810565318965085.34112291.9294147504.5165038412521187.69118.324.78532280264.67098993342886.8301.128101.35630.8313250.9463451680158.588158.44871.66642010360.90455337112.64439549108.4829.2221910088.5064.7764.9231.7340284.23101.9696.96742143.5012.08141.77130.204350609133.4743503624.525527.42126134.7400.22130.4231.000.48570235.47110.45563986.266756.982706.281.145712.32395.5541.52222976.53804.51446745.31364943.89177.9146.667506160.0140775875.195179.311095104.7484832.8291569.8797227.42275.0986533.33064298.29135.47916857432514.6674.59525.11.48678672.34510.51310.59193650001.69120257591.2151.591441038225057385837.231626.8609729302176220150.887074.57222202.262163987.012056776.098.353.816913.7948820.076849.81059.86744.303.18206314.22947.69474772538690532.3579.076.902.33299.0213.36383348.24256121.291302.6753.4474.797.22657138.37276.3652.3576.5252.24498.83890.9916.83237.305.04790.637323.27171.634.07978.6020.22395.194.75838.4621.70368.263.352329.2359.6308952070.2925613.299.13436.4111.51693.781370.4118120.385.871356.642.371659.0622193332222556249250694430112247481248985411.210889.19761170726676795936052.56420.38883780.470.4747.72754.03868.6754.38615.6355491021.6860413060.512.1011.9012.80116.77176.6335.60435.08416.091642668.642.93592.8511203.8318.394036.90160.97183.4204.8325.851646.26318.30046525.7768419723407234032062700008323.69402835.3228866.88328.704192.475929.564280.918468.8629040929.247180200001315000.02522516671637.72157.91322931806.874990017580731305.563119.56516.27716.35826.15928.617199224.62532192.658726.041489.70171.591469.864853.934458840910424850.7131.3023911.9793237.7030.143420.873122.2911.5970.3443.83311.86176.011944.435.70185645.951115517.4512.93786.7855.6674.718.26219.551107.474.5162962.714988108.21121.4957.79202116.215.4367583.7142.8395.583138.95123.891041161.51173.193.96661.888.715198.423.948881071.0217.52237.22246.28351.91528.22716.8224.383976.27689.53936.23978.51905.032421.962123.9941243.1003234.95680.74482.795452114.12453.306229.49428.254320.093148.28138.93185239268.441407.6472710592191.342274041.410308.3105.037332163882.02109.62236.28225.98717226527904315997733453729291.70279.62258.37197.3834.76327.412.66341378.25134.739868.470237.3277172160326.497820408660.527787870229489192.91793240343.999671222750016061844120318773650537068550488427805205130724693822343561378116670.70558980.9008894246.0565951157401824.3624.020.20939608201.6481115980669.8031.493271.81641.1143231.3969435597182.049111.62564.55707525730.51397428210.90645852111.2388.9895814130.3063.6863.7031.3243166.1087.8394.90057204.3422.81101.19131.868301941113.1673717296.068956.20525161.8440.3095.4434.780.27039294.9191.58869839.873004.590283.679.312512.61186.4522.28840212.65791.79311165.60450241.34273.1131.166945118.616277529.936140.89107676.39120690.43124133.88123864.16168.9813133.64958570.29928.13787646891890.0085.73942.31.29819769.69213.2962.208191250000.654251898113.5147.848594598313995420234.053629.3651534261590158942.176396.09148199.9192658380.052651410.8110.255.185705.1058920.770548.141613.17245.553.05135327.75152.22846332325530888.7597.7910.003.12313.0919.12251867.24639137.918413.4756.98105.205.32453187.78070.4485.1170.3185.29648.091465.7116.34366.705.321124.068717.76337.454.361372.1920.31589.775.021191.7221.40559.883.533338.3867.5585205330.3135169.317.75772.2111.911005.381630.4425279.976.171936.552.622254.3830303331841970236258920064211817775300361610.501395.22371445570008919502845.73314.76121580.640.6447.23343.98561.3440.42121.7757625696.1502683515.914.4015.8914.60227.20171.6040.00938.71922.570157419.430.549115.4825252.3525.534830.47180.87183.4248.7330.677766.92119.94467324.682120139328823286423015333311867568714.4541113.85469.545977.768480.446229.2612180.8042202851.977183066671909466.06510310002391.44227.45451972586.6571517255831067906.413140.08719.80719.94319.34921.046185029.53602259.009431.291508.63199.141273.944593.3450510713714156556.2836.7123131.9355343.5434.718591.94948.4312.3093.5645.9758.969101.613847.984.40832356.958110114.769.18399.2272.0797.417.16820.746131.052.7043063.855132114.08143.2774.02936124.524.4699792.5156.6755.526146.08125.34749166.31215.682.77155.899.525281.563.51065908.8220.31330.16336.78422.86568.40615.21422.7023113.7416.83825.361126.521730.879560.312944.5517297.1713825.431244.70916.46825373.58840.922291.79815.092540.892157.48140.76162697504.265529.4456122243237.236144386.15652.6136.585668905298.8573.36421.36415.39713314120803914264716614517214.00211.23195.64351.0625.50176.613.43791291.70519.155478.872295.3462459195406.896251824576.6962348881170368117.14635413782.2549707174843840603212009822552956035429817183427527546013220736913517240024389.87036642.5253170513.4476828953646655.50718.031.44425448341.304262780041042.410.9594411.03270.6330275.6257190648132.400205.20279.64579252451.20741448515.82031609147.4846.7822610028.7460.2360.2029.1726620.69115.6436.57260152.1791.46186.24156.906601201166.9705222103.455777.43468134.4460.17170.0025.700.38956174.06136.63149645.051927.063936.2109.2359.154584.5421.25322331.46773.69584423.92016834.88164.4191.206326210.1113765963.544723.731009136.8366910.9370866.4275062.51307.0078839.156514125.99541.708511135013148.8177.23500.01.71257583.98310.21615.38184263331.31921343073.7762.572651918172739730750.412619.8374960232899293059.022693.49699214.3751651621.671567216.546.532.950102.9195121.474646.56497.51363.303.78531264.14660.37583091550339482.9761.785.251.68393.1310.17500558.40408118.958232.0663.2963.156.78931147.27890.5044.1790.0244.40388.49697.6921.31187.286.60603.926027.23146.625.15773.9619.86301.936.19644.5022.01272.463.271804.9077.2346781510.3217920.7510.94364.2211.27531.481060.4114005.465.751040.152.981330.8115683331806410227202531463721781795186552115.558464.2921892176675133414664.14221.79942260.360.3649.32263.29858.4971.97912.2001372162.8214213288.49.599.1210.49417.60162.3830.18529.65212.334732626.853.88074.9311155.6414.357642.11150.51179.8160.2819.094477.35715.53835520.7296356617654176211564333336265.10307368.1920773.70248.853175.544487.283247.546403.8722079544.975461366671008567.36502723331243.88120.39207201364.323271013390538704.54097.60012.69612.84831.31935.707239419.29550147.476220.381168.35139.361206.086285.21462678688293341.1124.5928292.6063830.1024.224370.022825.318.8860.9037.62814.13358.075332.637.61730536.700129815.6116.82780.5843.5763.7610.02416.54386.794.4940647.17516084.3698.4507.9345588.446.7986218.2119.7605.694105.23101.801380126.08135.995.17852.366.838173.335.552541073.6167.39208.12213.44279.57449.56618.85127.827176.438214.68040.48823.26710.789324.462040.3547204.5984740.33481.39395.436606130.80379.719168.39345.346267.077106.01117.99227813204.015398.7674579896154.852363062.613432.7101.027603185067.88131.34168.04184.51319719434649519628345542560291.26293.87265.18139.5740.49409.892.41341418.46749.6021240.060232.6778600819316.268018926359.617915944138418392.27799996115.63127857290350211726656087246764339927906864501031066335681332913806810582526262865.10655470.62812732948.4994566629443540.8722.515.68351022153.648652145533566.8501.765882.56361.5706227.3470817373209.22485.30864.52708964690.36013872548.5635840590.642611.038015940.6165.4965.2432.1448014.8683.7165.01824199.2544.1572.7878.37873794781.8791574288.498697.12100140.3520.4067.9940.370.75768396.8574.06190314.096269.1118066.767.757814.76647.4942.68247856.87746.242151710.59165146.20248.0104.32436187.702089760.758364.07112454.64150874.91158717.74160748.18141.456429.68921557.47822.97165469041355.1047.02970.71.11552895.93012.9694.515191190672.911831319135.7831.121479799437673908928.803734.7164401421140115534.640048.66832198.0563693215.303756154.9815.077.307487.1447018.819653.149718.61757.652.67129374.22928.06273272018442961.2605.1513.194.55312.5225.54174607.19005139.017574.8857.68138.505.43950183.82283.3095.9583.1896.10928.401882.3415.45516.905.101560.9911518.32435.774.281863.3620.57776.845.011588.3021.85731.483.514486.5836.3144335380.346506.757.891011.6112.041326.242430.4333111.826.212563.792.513099.9038250002706385268325132843992292956143345938.30854120.35117090800013000673737.10713.82281680.840.8447.57340.09968.8028.53329.1655923331.4847683350.920.6522.1116.82034.02176.6947.05345.38332.177580769.425.249139.1124785.6534.236529.502100.05183.6315.4548.8211220.52621.72388626.451155044419614188928256333315777749237.4156703.06617.797928.20112078225.2616010.7455781750.847726166672511125.72495846673149.43298.15600673424.5793970336931404677.296162.00225.90726.40817.72917.666171540.15571365.317243.892051.60239.931592.333843.1130716720117712574.9651.6321271.1667458.2544.598830.776603.6716.43107.5348.8608.373143.302059.483.52156569.960105115.127.030109.5797.77114.886.17422.174162.472.8908780.523978154.02166.3744.24532168.133.56813807.2175.2055.480187.73169.81620224.02297.032.11760.1411.766357.602.70782907.5271.84410.42419.93554.16631.94214.58550.2806139.06310.39335.01936.34976.202403.462407.8998254.7731634.79697.15524.511446109.73491.502202.27464.506339.220126.03118.52191756291.509421.8370226868177.016270736.09585.5108.817054497975.83109.36242.01246.66316962527391416831329772183259.35256.89233.74199.3134.68315.302.67554373.66834.649866.228244.4474723579333.417623218562.307567015127516795.09774340113.969272516116715131382442717601876559365832975290781478597023876211613586378371975.68023170.9618324156.2380054316120272.10320.020.09039830202.530906112383707.6311.413491.83271.1237231.5568771814193.968120.85866.49687359850.50599067610.71446671108.9489.1834011606.2265.8265.8732.4041255.9890.0305.93207168.6162.77103.25104.124424270103.7308204586.254187.09240140.9790.3097.8434.720.60813305.2289.62571721.175445.792280.778.392612.76246.4902.12337547.76812.29329687.18706239.53221.2129.085729120.9147937389.246270.82113178.09108719.12116236.16118177.50210.3926332.84655475.25031.85585619371973.9969.98798.41.38792719.98511.51036.15180782002.205601831110.5239.010788519308855372633.141830.1852528721554156249.179276.31605197.9752762180.282749241.8010.415.347535.2733320.184049.548113.29535.552.96692336.91139.28207429124972745.5647.109.243.12331.3718.09247198.19253121.994413.4858.56102.356.41451155.87781.5673.5180.7974.22679.281274.2717.05351.235.641059.998919.68304.224.641286.3122.11541.925.331120.7723.47510.623.643221.2050.6996218760.3233784.748.78681.1312.82934.241810.4623774.886.491835.782.72184.942807000216669925828494750279227855631309349.74591102.6001363440009172694244.34217.29114790.640.6450.45845.58656.9639.64521.2409648144.0033573338.214.6615.7114.53627.43177.9341.77040.17721.989656873.233.029111.9917215.9225.008632.55680.93183.6256.6534.598864.96220.21062286.195112461311193144023359666711623539442.2441241.45453.875850.598184.445939.5211892.6939475809.787284600001758023.97531470002229.57210.64425072468.4367820249371008536.526141.76820.52820.74320.25422.123191429.82526258.378031.671542.11204.701351.944413.5756811850713993557.6837.2223741.9736244.4635.234566.424438.0112.7092.0746.1569.414101.800249.244.44790457.233110914.769.943103.2571.6497.277.21521.505130.383.8069366.475037117.12145.0426.46357127.694.43710640.8162.4245.815150.51131.13781172.91216.673.00454.499.266284.343.454611034.9231.05331.83337.46428.68581.67215.05535.0223107.7410OpenBenchmarking.org

LAMMPS Molecular Dynamics Simulator

LAMMPS is a classical molecular dynamics code, and an acronym for Large-scale Atomic/Molecular Massively Parallel Simulator. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 23Jun2022Model: 20k AtomsEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248848121620SE +/- 0.191, N = 9SE +/- 0.298, N = 6SE +/- 0.014, N = 3SE +/- 0.108, N = 9SE +/- 0.004, N = 3SE +/- 0.174, N = 9SE +/- 0.297, N = 9SE +/- 0.177, N = 910.39314.6806.8389.5398.9518.21513.6917.5791. (CXX) g++ options: -O3 -lm -ldl

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 512 - Model: ResNet-50EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488918273645SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.31, N = 8SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 335.0140.4825.3636.2330.4930.7041.0222.98

OpenRadioss

OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: Chrysler Neon 1MEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24882004006008001000SE +/- 9.60, N = 3SE +/- 8.46, N = 9SE +/- 2.88, N = 3SE +/- 2.98, N = 3SE +/- 1.89, N = 3SE +/- 10.68, N = 9SE +/- 4.49, N = 3SE +/- 17.37, N = 7936.34823.261126.52978.51967.201003.59891.84983.54

Timed Linux Kernel Compilation

This test times how long it takes to build the Linux kernel in a default configuration (defconfig) for the architecture being tested or alternatively an allmodconfig for building all possible kernel modules for the build. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.8Build: allmodconfigEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488400800120016002000SE +/- 1.21, N = 3SE +/- 0.27, N = 3SE +/- 0.90, N = 3SE +/- 0.76, N = 3SE +/- 0.16, N = 3SE +/- 0.32, N = 3SE +/- 0.38, N = 3SE +/- 0.63, N = 3976.20710.791730.88905.031319.931310.96714.381508.04

OpenRadioss

OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: INIVOL and Fluid Structure Interaction Drop ContainerEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488130260390520650SE +/- 9.39, N = 6SE +/- 4.35, N = 9SE +/- 0.28, N = 3SE +/- 4.68, N = 5SE +/- 4.81, N = 6SE +/- 5.87, N = 9SE +/- 2.52, N = 3SE +/- 8.95, N = 9403.46324.46560.31421.96494.79585.75322.67538.42

OpenFOAM

OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Medium Mesh Size - Execution TimeEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248860012001800240030002407.902040.352944.552123.992781.242827.882351.512260.991. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfoamToVTK -ldynamicMesh -llagrangian -lgenericPatchFields -lfileFormats -lOpenFOAM -ldl -lm

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Medium Mesh Size - Mesh TimeEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248860120180240300254.77204.60297.17243.10262.84296.21226.01225.951. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfoamToVTK -ldynamicMesh -llagrangian -lgenericPatchFields -lfileFormats -lOpenFOAM -ldl -lm

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 256 - Model: ResNet-50EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488918273645SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 334.7940.3325.4334.9530.5030.6340.6823.07

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Barbershop - Compute: CPU-OnlyEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248830060090012001500SE +/- 0.89, N = 3SE +/- 0.24, N = 3SE +/- 0.31, N = 3SE +/- 0.50, N = 3SE +/- 1.01, N = 3SE +/- 0.87, N = 3SE +/- 0.46, N = 3SE +/- 4.06, N = 3697.15481.391244.70680.74942.29925.38481.151281.54

Timed Node.js Compilation

This test profile times how long it takes to build/compile Node.js itself from source. Node.js is a JavaScript run-time built from the Chrome V8 JavaScript engine while itself is written in C/C++. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Node.js Compilation 21.7.2Time To CompileEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24882004006008001000SE +/- 0.29, N = 3SE +/- 0.55, N = 3SE +/- 0.13, N = 3SE +/- 0.11, N = 3SE +/- 0.61, N = 3SE +/- 0.38, N = 3SE +/- 0.68, N = 3SE +/- 1.08, N = 3524.51395.44916.47482.80743.11739.34421.43915.94

OpenVKL

OpenVKL is the Intel Open Volume Kernel Library that offers high-performance volume computation kernels and part of the Intel oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgItems / Sec, More Is BetterOpenVKL 2.0.0Benchmark: vklBenchmarkCPU ISPCEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488130260390520650SE +/- 0.67, N = 3SE +/- 0.33, N = 3SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.67, N = 3SE +/- 0.33, N = 3446606253452327331603288MIN: 34 / MAX: 6230MIN: 47 / MAX: 8240MIN: 19 / MAX: 3690MIN: 34 / MAX: 5835MIN: 24 / MAX: 4788MIN: 25 / MAX: 4820MIN: 48 / MAX: 8450MIN: 20 / MAX: 4338

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 512 - Model: GoogLeNetEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.05, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.00, N = 3109.73130.8073.58114.1290.1391.90132.6276.13

Timed LLVM Compilation

This test times how long it takes to compile/build the LLVM compiler stack. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 16.0Build System: Unix MakefilesEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24882004006008001000SE +/- 0.73, N = 3SE +/- 2.07, N = 3SE +/- 0.63, N = 3SE +/- 0.93, N = 3SE +/- 0.24, N = 3SE +/- 0.38, N = 3SE +/- 1.00, N = 3SE +/- 0.50, N = 3491.50379.72840.92453.31669.64662.52387.01719.39

OpenRadioss

OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: Bird Strike on WindshieldEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248860120180240300SE +/- 4.67, N = 12SE +/- 3.16, N = 9SE +/- 0.57, N = 3SE +/- 1.41, N = 3SE +/- 0.53, N = 3SE +/- 7.66, N = 9SE +/- 2.93, N = 12SE +/- 3.59, N = 9202.27168.39291.79229.49238.13266.65183.64259.88

Timed LLVM Compilation

This test times how long it takes to compile/build the LLVM compiler stack. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 16.0Build System: NinjaEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24882004006008001000SE +/- 0.48, N = 3SE +/- 0.47, N = 3SE +/- 0.14, N = 3SE +/- 0.34, N = 3SE +/- 0.10, N = 3SE +/- 0.77, N = 3SE +/- 0.24, N = 3SE +/- 0.51, N = 3464.51345.35815.09428.25644.77636.83354.76697.94

Timed Gem5 Compilation

This test times how long it takes to compile Gem5. Gem5 is a simulator for computer system architecture research. Gem5 is widely used for computer architecture research within the industry, academia, and more. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Gem5 Compilation 23.0.1Time To CompileEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488120240360480600SE +/- 1.21, N = 3SE +/- 3.30, N = 4SE +/- 0.50, N = 3SE +/- 1.50, N = 3SE +/- 0.25, N = 3SE +/- 0.34, N = 3SE +/- 0.80, N = 3SE +/- 0.89, N = 3339.22267.08540.89320.09438.62432.22273.49468.38

OpenRadioss

OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: Rubber O-Ring Seal InstallationEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24884080120160200SE +/- 7.92, N = 9SE +/- 3.75, N = 15SE +/- 0.54, N = 3SE +/- 3.80, N = 9SE +/- 0.24, N = 3SE +/- 8.68, N = 9SE +/- 1.00, N = 3SE +/- 5.45, N = 12126.03106.01157.48148.28127.32162.03110.05177.74

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: Bumper BeamEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 2.01, N = 15SE +/- 1.50, N = 3SE +/- 0.20, N = 3SE +/- 1.43, N = 5SE +/- 0.23, N = 3SE +/- 3.91, N = 12SE +/- 1.49, N = 15SE +/- 2.49, N = 12118.52117.99140.76138.93124.70145.32108.64153.90

MariaDB

This is a MariaDB MySQL database server benchmark making use of sysbench rather than the existing pts/mysqlslap test profile that uses MariaDB with mysqlslap/mariadb-slap as the benchmark driver. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.5Test: oltp_update_non_index - Threads: 64EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248850K100K150K200K250KSE +/- 159.15, N = 3SE +/- 355.20, N = 3SE +/- 223.22, N = 3SE +/- 1725.95, N = 9SE +/- 120.42, N = 3SE +/- 295.88, N = 3SE +/- 290.99, N = 3SE +/- 516.32, N = 3191756227813162697185239197477201401233313220068-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto1. (CXX) g++ options: -fPIC -pie -fstack-protector -O3 -lpthread -lz -ldl -lm

Helsing

Helsing is an open-source POSIX vampire number generator. This test profile measures the time it takes to generate vampire numbers between varying numbers of digits. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterHelsing 1.0-betaDigit Range: 14 digitEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488110220330440550SE +/- 0.29, N = 3SE +/- 0.19, N = 3SE +/- 0.52, N = 3SE +/- 0.10, N = 3SE +/- 0.31, N = 3SE +/- 0.18, N = 3SE +/- 0.37, N = 3SE +/- 0.23, N = 3291.51204.02504.27268.44386.16375.92200.77437.411. (CC) gcc options: -O2 -pthread

Apache IoTDB

Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgAverage Latency, Fewer Is BetterApache IoTDB 1.2Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 400EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488110220330440550SE +/- 4.87, N = 3SE +/- 1.42, N = 3SE +/- 8.54, N = 5SE +/- 1.07, N = 3SE +/- 3.04, N = 3SE +/- 4.56, N = 3SE +/- 4.31, N = 3SE +/- 1.92, N = 3421.83398.76529.44407.64465.51462.14399.56524.13MAX: 27011.65MAX: 26832.58MAX: 28194.55MAX: 26798.87MAX: 27554.23MAX: 27920.53MAX: 27980.82MAX: 26588.63

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.2Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 400EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248816M32M48M64M80MSE +/- 367807.54, N = 3SE +/- 219419.50, N = 3SE +/- 580368.90, N = 5SE +/- 171497.11, N = 3SE +/- 178945.47, N = 3SE +/- 441409.11, N = 3SE +/- 597457.28, N = 3SE +/- 207001.08, N = 37022686874579896561222437271059263954176641825637371602556494499

GPAW

GPAW is a density-functional theory (DFT) Python code based on the projector-augmented wave (PAW) method and the atomic simulation environment (ASE). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterGPAW 23.6Input: Carbon NanotubeEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248850100150200250SE +/- 2.27, N = 3SE +/- 3.67, N = 10SE +/- 0.19, N = 3SE +/- 1.91, N = 3SE +/- 0.06, N = 3SE +/- 2.92, N = 3SE +/- 1.88, N = 3SE +/- 8.24, N = 12177.02154.85237.24191.34203.43205.87154.99228.901. (CC) gcc options: -shared -fwrapv -O2 -lxc -lblas -lmpi

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgverify/s, More Is BetterOpenSSL 3.3Algorithm: RSA4096EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248880K160K240K320K400KSE +/- 313.76, N = 3SE +/- 494.21, N = 3SE +/- 34.43, N = 3SE +/- 923.26, N = 3SE +/- 13.93, N = 3SE +/- 20.61, N = 3SE +/- 143.01, N = 3SE +/- 1626.68, N = 3270736.0363062.6144386.1274041.4190537.1194210.3373632.3195618.81. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenBenchmarking.orgsign/s, More Is BetterOpenSSL 3.3Algorithm: RSA4096EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24883K6K9K12K15KSE +/- 3.17, N = 3SE +/- 2.41, N = 3SE +/- 3.91, N = 3SE +/- 14.97, N = 3SE +/- 2.72, N = 3SE +/- 2.36, N = 3SE +/- 5.90, N = 3SE +/- 28.09, N = 39585.513432.75652.610308.37335.27697.314289.73000.21. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

Apache IoTDB

Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgAverage Latency, Fewer Is BetterApache IoTDB 1.2Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 100EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.12, N = 3SE +/- 0.15, N = 3SE +/- 0.11, N = 3SE +/- 0.27, N = 3SE +/- 0.64, N = 3SE +/- 0.48, N = 3SE +/- 0.21, N = 3SE +/- 0.51, N = 3108.81101.02136.58105.03122.33122.55103.37137.92MAX: 23915.13MAX: 23859.84MAX: 23944.74MAX: 23941.39MAX: 23948.29MAX: 23938.16MAX: 23852.59MAX: 23917.2

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.2Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 100EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248816M32M48M64M80MSE +/- 59581.06, N = 3SE +/- 138400.03, N = 3SE +/- 45916.59, N = 3SE +/- 81989.58, N = 3SE +/- 302032.90, N = 3SE +/- 268932.10, N = 3SE +/- 205604.00, N = 3SE +/- 209117.79, N = 37054497976031850566890527332163863086247629514587407524456035278

OpenRadioss

OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: Cell Phone Drop TestEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820406080100SE +/- 1.62, N = 15SE +/- 0.42, N = 3SE +/- 0.21, N = 3SE +/- 1.45, N = 15SE +/- 0.37, N = 3SE +/- 3.06, N = 15SE +/- 1.68, N = 12SE +/- 3.30, N = 1575.8367.8898.8582.0280.0495.4867.7193.26

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 256 - Model: GoogLeNetEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.04, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.08, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.08, N = 3SE +/- 0.05, N = 3109.36131.3473.36109.6290.1991.77132.1376.84

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Pabellon Barcelona - Compute: CPU-OnlyEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248890180270360450SE +/- 0.08, N = 3SE +/- 0.14, N = 3SE +/- 0.44, N = 3SE +/- 0.05, N = 3SE +/- 0.34, N = 3SE +/- 0.40, N = 3SE +/- 0.18, N = 3SE +/- 1.05, N = 3242.01168.04421.36236.28323.68315.13165.35411.15

Timed Godot Game Engine Compilation

This test times how long it takes to compile the Godot Game Engine. Godot is a popular, open-source, cross-platform 2D/3D game engine and is built using the SCons build system and targeting the X11 platform. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Godot Game Engine Compilation 4.0Time To CompileEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248890180270360450SE +/- 0.49, N = 3SE +/- 0.60, N = 3SE +/- 0.05, N = 3SE +/- 0.51, N = 3SE +/- 0.16, N = 3SE +/- 0.15, N = 3SE +/- 0.40, N = 3SE +/- 0.34, N = 3246.66184.51415.40225.99330.82326.16188.53364.42

MariaDB

This is a MariaDB MySQL database server benchmark making use of sysbench rather than the existing pts/mysqlslap test profile that uses MariaDB with mysqlslap/mariadb-slap as the benchmark driver. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.5Test: oltp_read_write - Threads: 64EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248840K80K120K160K200KSE +/- 99.82, N = 3SE +/- 91.63, N = 3SE +/- 193.47, N = 3SE +/- 296.87, N = 3SE +/- 284.19, N = 3SE +/- 211.83, N = 3SE +/- 389.87, N = 3SE +/- 562.40, N = 3169625197194133141172265166060168593201182186418-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto1. (CXX) g++ options: -fPIC -pie -fstack-protector -O3 -lpthread -lz -ldl -lm

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.5Test: oltp_write_only - Threads: 64EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248870K140K210K280K350KSE +/- 431.10, N = 3SE +/- 391.18, N = 3SE +/- 493.89, N = 3SE +/- 971.36, N = 3SE +/- 204.75, N = 3SE +/- 419.19, N = 3SE +/- 56.85, N = 3SE +/- 798.38, N = 3273914346495208039279043260139265119348099296034-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto1. (CXX) g++ options: -fPIC -pie -fstack-protector -O3 -lpthread -lz -ldl -lm

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.5Test: oltp_update_index - Threads: 64EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248840K80K120K160K200KSE +/- 411.40, N = 3SE +/- 412.73, N = 3SE +/- 254.61, N = 3SE +/- 173.91, N = 3SE +/- 133.22, N = 3SE +/- 308.54, N = 3SE +/- 282.52, N = 3SE +/- 250.58, N = 3168313196283142647159977173080177579198538187089-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto1. (CXX) g++ options: -fPIC -pie -fstack-protector -O3 -lpthread -lz -ldl -lm

Stockfish

This is a test of Stockfish, an advanced open-source C++11 chess benchmark that can scale up to 1024 CPU threads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNodes Per Second, More Is BetterStockfish 16.1Chess BenchmarkEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248810M20M30M40M50MSE +/- 353999.00, N = 15SE +/- 611568.23, N = 15SE +/- 169306.09, N = 3SE +/- 379095.21, N = 4SE +/- 191557.41, N = 3SE +/- 160295.89, N = 3SE +/- 574948.55, N = 15SE +/- 121681.53, N = 152977218345542560166145173345372921844343231438584427085816898291-mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl-mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl-mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl-mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl-mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl-mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl-mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl1. (CXX) g++ options: -lgcov -m64 -lpthread -fno-exceptions -std=c++17 -fno-peel-loops -fno-tracer -pedantic -O3 -funroll-loops -msse -msse3 -mpopcnt -mavx2 -mbmi -msse4.1 -mssse3 -msse2 -mbmi2 -flto -flto-partition=one -flto=jobserver

ClickHouse

ClickHouse is an open-source, high performance OLAP data management system. This test profile uses ClickHouse's standard benchmark recommendations per https://clickhouse.com/docs/en/operations/performance-test/ / https://github.com/ClickHouse/ClickBench/tree/main/clickhouse with the 100 million rows web analytics dataset. The reported value is the query processing time using the geometric mean of all separate queries performed as an aggregate. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Minute, Geo Mean, More Is BetterClickHouse 22.12.3.5100M Rows Hits Dataset, Third RunEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248860120180240300SE +/- 4.31, N = 3SE +/- 2.43, N = 3SE +/- 1.41, N = 3SE +/- 2.40, N = 3SE +/- 2.33, N = 3SE +/- 4.93, N = 3SE +/- 2.29, N = 5SE +/- 1.74, N = 3259.35291.26214.00291.70233.30232.47251.80249.90MIN: 13.75 / MAX: 10000MIN: 16.56 / MAX: 7500MIN: 7.95 / MAX: 10000MIN: 14.85 / MAX: 10000MIN: 10.46 / MAX: 10000MIN: 10.67 / MAX: 12000MIN: 15.75 / MAX: 10000MIN: 9.25 / MAX: 12000

OpenBenchmarking.orgQueries Per Minute, Geo Mean, More Is BetterClickHouse 22.12.3.5100M Rows Hits Dataset, Second RunEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248860120180240300SE +/- 2.91, N = 3SE +/- 3.27, N = 3SE +/- 1.04, N = 3SE +/- 1.95, N = 3SE +/- 0.75, N = 3SE +/- 1.00, N = 3SE +/- 1.51, N = 5SE +/- 3.87, N = 3256.89293.87211.23279.62230.37236.47248.89249.50MIN: 13.95 / MAX: 10000MIN: 18.95 / MAX: 10000MIN: 7.98 / MAX: 10000MIN: 14.87 / MAX: 8571.43MIN: 10.32 / MAX: 10000MIN: 10.66 / MAX: 10000MIN: 15.92 / MAX: 8571.43MIN: 9.27 / MAX: 12000

OpenBenchmarking.orgQueries Per Minute, Geo Mean, More Is BetterClickHouse 22.12.3.5100M Rows Hits Dataset, First Run / Cold CacheEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248860120180240300SE +/- 0.94, N = 3SE +/- 1.16, N = 3SE +/- 1.29, N = 3SE +/- 1.12, N = 3SE +/- 1.31, N = 3SE +/- 1.84, N = 3SE +/- 2.37, N = 5SE +/- 1.98, N = 3233.74265.18195.64258.37213.47222.47227.39234.39MIN: 13.67 / MAX: 10000MIN: 14.2 / MAX: 10000MIN: 7.8 / MAX: 10000MIN: 14.45 / MAX: 10000MIN: 10.21 / MAX: 8571.43MIN: 10.49 / MAX: 12000MIN: 13.31 / MAX: 10000MIN: 9.14 / MAX: 12000

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Classroom - Compute: CPU-OnlyEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248880160240320400SE +/- 0.27, N = 3SE +/- 0.33, N = 3SE +/- 0.40, N = 3SE +/- 0.13, N = 3SE +/- 0.28, N = 3SE +/- 0.07, N = 3SE +/- 0.07, N = 3SE +/- 1.57, N = 3199.31139.57351.06197.38267.86259.46136.11336.94

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 64 - Model: ResNet-50EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488918273645SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.03, N = 334.6840.4925.5034.7630.6430.5340.2423.10

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 512 - Model: AlexNetEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248890180270360450SE +/- 0.04, N = 3SE +/- 0.06, N = 3SE +/- 0.09, N = 3SE +/- 0.08, N = 3SE +/- 0.04, N = 3SE +/- 0.55, N = 3SE +/- 1.74, N = 3SE +/- 0.02, N = 3315.30409.89176.61327.41238.89240.23425.47192.28

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: ResNet50 v1-12-int8 - Device: CPU - Executor: StandardEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881.00392.00783.01174.01565.0195SE +/- 0.00434, N = 3SE +/- 0.06313, N = 15SE +/- 0.05218, N = 15SE +/- 0.05991, N = 15SE +/- 0.03376, N = 15SE +/- 0.00310, N = 3SE +/- 0.05300, N = 15SE +/- 0.03583, N = 32.675542.413413.437912.663412.918463.091502.382994.461701. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: ResNet50 v1-12-int8 - Device: CPU - Executor: StandardEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248890180270360450SE +/- 0.61, N = 3SE +/- 11.38, N = 15SE +/- 4.18, N = 15SE +/- 9.13, N = 15SE +/- 3.81, N = 15SE +/- 0.32, N = 3SE +/- 9.29, N = 15SE +/- 1.81, N = 3373.67418.47291.71378.25343.18323.38422.42224.111. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

miniBUDE

MiniBUDE is a mini application for the the core computation of the Bristol University Docking Engine (BUDE). This test profile currently makes use of the OpenMP implementation of miniBUDE for CPU benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBillion Interactions/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM2EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881122334455SE +/- 0.02, N = 3SE +/- 0.09, N = 3SE +/- 0.00, N = 3SE +/- 0.03, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.06, N = 3SE +/- 0.01, N = 334.6549.6019.1634.7425.8926.3649.9912.551. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

OpenBenchmarking.orgGFInst/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM2EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248830060090012001500SE +/- 0.58, N = 3SE +/- 2.18, N = 3SE +/- 0.06, N = 3SE +/- 0.75, N = 3SE +/- 0.07, N = 3SE +/- 0.13, N = 3SE +/- 1.54, N = 3SE +/- 0.15, N = 3866.231240.06478.87868.47647.13658.931249.62313.741. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

Apache IoTDB

Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgAverage Latency, Fewer Is BetterApache IoTDB 1.2Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 400EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248860120180240300SE +/- 3.20, N = 3SE +/- 1.32, N = 3SE +/- 1.17, N = 3SE +/- 0.75, N = 3SE +/- 6.43, N = 3SE +/- 1.82, N = 3SE +/- 2.35, N = 3SE +/- 4.53, N = 3244.44232.67295.34237.32260.60268.42235.16290.14MAX: 26631.84MAX: 26680.29MAX: 26742.14MAX: 26320.16MAX: 26603.07MAX: 27440.82MAX: 26882.16MAX: 26535.45

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.2Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 400EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820M40M60M80M100MSE +/- 718192.76, N = 3SE +/- 285870.90, N = 3SE +/- 77022.60, N = 3SE +/- 169341.63, N = 3SE +/- 676495.92, N = 3SE +/- 261270.67, N = 3SE +/- 360981.97, N = 3SE +/- 787042.09, N = 37472357978600819624591957717216070107030691441647765891462655549

OpenBenchmarking.orgAverage Latency, Fewer Is BetterApache IoTDB 1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 400EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248890180270360450SE +/- 1.97, N = 3SE +/- 1.19, N = 3SE +/- 2.59, N = 3SE +/- 5.21, N = 3SE +/- 6.58, N = 3SE +/- 6.12, N = 3SE +/- 5.40, N = 3SE +/- 10.20, N = 3333.41316.26406.89326.49350.00347.97321.29382.48MAX: 26859.8MAX: 27135.69MAX: 27507.51MAX: 27383.3MAX: 27235.37MAX: 27085.8MAX: 27126.33MAX: 27376.22

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 400EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820M40M60M80M100MSE +/- 191894.24, N = 3SE +/- 243124.87, N = 3SE +/- 173074.65, N = 3SE +/- 707677.76, N = 3SE +/- 612351.99, N = 3SE +/- 255843.12, N = 3SE +/- 818556.17, N = 3SE +/- 922667.78, N = 37623218580189263625182457820408671203893713387407845431664886364

OpenBenchmarking.orgAverage Latency, Fewer Is BetterApache IoTDB 1.2Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 100EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820406080100SE +/- 0.07, N = 3SE +/- 0.05, N = 3SE +/- 0.12, N = 3SE +/- 0.13, N = 3SE +/- 0.05, N = 3SE +/- 0.29, N = 3SE +/- 0.29, N = 3SE +/- 0.20, N = 362.3059.6176.6960.5268.5868.1660.6677.24MAX: 23874.13MAX: 23860.38MAX: 23999.1MAX: 23884.06MAX: 23897.06MAX: 23889.9MAX: 23861.25MAX: 23891.04

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.2Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 100EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820M40M60M80M100MSE +/- 132686.57, N = 3SE +/- 118053.95, N = 3SE +/- 57777.29, N = 3SE +/- 198325.80, N = 3SE +/- 77775.76, N = 3SE +/- 296062.71, N = 3SE +/- 390613.00, N = 3SE +/- 232749.13, N = 37567015179159441623488817787870269342377695614497760562761842388

BRL-CAD

BRL-CAD is a cross-platform, open-source solid modeling system with built-in benchmark mode. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgVGR Performance Metric, More Is BetterBRL-CAD 7.38.2VGR Performance MetricEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248890K180K270K360K450K2751673841831703682948912217672261983971381941271. (CXX) g++ options: -std=c++17 -pipe -fvisibility=hidden -fno-strict-aliasing -fno-common -fexceptions -ftemplate-depth-128 -m64 -ggdb3 -O3 -fipa-pta -fstrength-reduce -finline-functions -flto -ltcl8.6 -lnetpbm -lregex_brl -lz_brl -lassimp -ldl -lm -ltk8.6

Apache IoTDB

Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgAverage Latency, Fewer Is BetterApache IoTDB 1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 100EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.32, N = 3SE +/- 0.56, N = 3SE +/- 0.72, N = 3SE +/- 0.40, N = 3SE +/- 0.39, N = 3SE +/- 0.24, N = 3SE +/- 0.82, N = 3SE +/- 0.54, N = 395.0992.27117.1492.91105.70105.5892.08119.49MAX: 12619.54MAX: 11332.45MAX: 15130.91MAX: 12607.68MAX: 13869.85MAX: 15103.61MAX: 11356.99MAX: 13845.12

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 100EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820M40M60M80M100MSE +/- 105860.93, N = 3SE +/- 356287.02, N = 3SE +/- 200521.84, N = 3SE +/- 51778.51, N = 3SE +/- 216731.24, N = 3SE +/- 282897.65, N = 3SE +/- 368630.91, N = 3SE +/- 262085.32, N = 37743401179999611635413787932403470293571703052857931531962558098

SVT-HEVC

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 1 - Input: Bosphorus 4KEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881.3052.613.9155.226.525SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.03, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 33.965.632.253.992.943.045.802.261. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: ChaCha20EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248830000M60000M90000M120000M150000MSE +/- 14660801.77, N = 3SE +/- 45400804.32, N = 3SE +/- 41193652.45, N = 3SE +/- 26095977.93, N = 3SE +/- 10772886.17, N = 3SE +/- 32786296.71, N = 3SE +/- 13732325.68, N = 3SE +/- 7700475.42, N = 39272516116712785729035049707174843967122275006657277460768366108810129853526473346518558001. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: AES-256-GCMEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248850000M100000M150000M200000M250000MSE +/- 10292339.65, N = 3SE +/- 124036322.98, N = 3SE +/- 23525395.17, N = 3SE +/- 51787792.79, N = 3SE +/- 28146850.41, N = 3SE +/- 101608115.02, N = 3SE +/- 27189836.88, N = 3SE +/- 5611535.27, N = 315131382442721172665608784060321200160618441203110425587253113941258877215336847677561932491901. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: AES-128-GCMEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248850000M100000M150000M200000M250000MSE +/- 82240254.30, N = 3SE +/- 834503022.99, N = 3SE +/- 30766844.46, N = 3SE +/- 100195492.71, N = 3SE +/- 18100529.72, N = 3SE +/- 48206635.82, N = 3SE +/- 197196146.33, N = 3SE +/- 45687236.99, N = 317601876559324676433992798225529560187736505370129046626940133134074673251224402513648127256801. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: ChaCha20-Poly1305EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820000M40000M60000M80000M100000MSE +/- 17029222.81, N = 3SE +/- 37115945.76, N = 3SE +/- 14961656.63, N = 3SE +/- 21541900.76, N = 3SE +/- 18319786.28, N = 3SE +/- 11858939.41, N = 3SE +/- 63328398.60, N = 3SE +/- 5622113.45, N = 365832975290906864501033542981718368550488427472542301734854273966392129622150238002630201. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: SHA512EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24882000M4000M6000M8000M10000MSE +/- 1960683.64, N = 3SE +/- 5901271.36, N = 3SE +/- 432194.03, N = 3SE +/- 7000138.46, N = 3SE +/- 485369.37, N = 3SE +/- 749962.39, N = 3SE +/- 2951460.39, N = 3SE +/- 2787035.99, N = 378147859701066335681342752754608052051307564435758357944053071104157992344293971601. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: SHA256EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24887000M14000M21000M28000M35000MSE +/- 5636428.54, N = 3SE +/- 8957910.25, N = 3SE +/- 4935253.44, N = 3SE +/- 9363717.55, N = 3SE +/- 7769960.49, N = 3SE +/- 8102567.82, N = 3SE +/- 7892274.19, N = 3SE +/- 27097230.48, N = 323876211613329138068101322073691324693822343174027128101787635833733989652273123365572631. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

MariaDB mariadb-slap

This is a MariaDB MySQL database server benchmark making use of mysqlslap (mariadb-slap). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB mariadb-slap 11.5Clients: 128EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488130260390520650SE +/- 0.23, N = 3SE +/- 0.71, N = 3SE +/- 0.54, N = 3SE +/- 0.93, N = 3SE +/- 0.66, N = 3SE +/- 0.74, N = 3SE +/- 0.54, N = 3SE +/- 1.05, N = 3586582517561565580612574-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto1. (CXX) g++ options: -fPIC -pie -fstack-protector -O3 -lpthread -lz -ldl -lm

Speedb

Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Read While WritingEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881.1M2.2M3.3M4.4M5.5MSE +/- 18321.18, N = 3SE +/- 35376.76, N = 15SE +/- 17743.52, N = 15SE +/- 55041.91, N = 13SE +/- 24740.75, N = 15SE +/- 14207.26, N = 3SE +/- 26941.48, N = 3SE +/- 20209.24, N = 3378371952626282400243378116631896503194971511989125488371. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

Xcompact3d Incompact3d

Xcompact3d Incompact3d is a Fortran-MPI based, finite difference high-performance code for solving the incompressible Navier-Stokes equation and as many as you need scalar transport equations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterXcompact3d Incompact3d 2021-03-11Input: input.i3d 193 Cells Per DirectionEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820406080100SE +/- 0.40, N = 3SE +/- 0.83, N = 3SE +/- 0.07, N = 3SE +/- 0.60, N = 15SE +/- 0.30, N = 3SE +/- 0.27, N = 3SE +/- 0.58, N = 10SE +/- 0.87, N = 1575.6865.1189.8770.7185.3487.7773.8873.721. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

PostgreSQL

This is a benchmark of PostgreSQL using the integrated pgbench for facilitating the database benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 800 - Mode: Read Only - Average LatencyEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24880.56811.13621.70432.27242.8405SE +/- 0.007, N = 3SE +/- 0.002, N = 3SE +/- 0.021, N = 9SE +/- 0.012, N = 3SE +/- 0.010, N = 3SE +/- 0.023, N = 3SE +/- 0.006, N = 3SE +/- 0.005, N = 30.9610.6282.5250.9001.9292.0510.7221.8521. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 800 - Mode: Read OnlyEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488300K600K900K1200K1500KSE +/- 6188.19, N = 3SE +/- 4344.86, N = 3SE +/- 2616.05, N = 9SE +/- 11913.26, N = 3SE +/- 2047.43, N = 3SE +/- 4300.20, N = 3SE +/- 8757.92, N = 3SE +/- 1154.98, N = 3832415127329431705188942441475039001811082434319621. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OSPRay

Intel OSPRay is a portable ray-tracing engine for high-performance, high-fidelity scientific visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 3.1Benchmark: particle_volume/scivis/real_timeEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24883691215SE +/- 0.00310, N = 3SE +/- 0.00215, N = 3SE +/- 0.00497, N = 3SE +/- 0.00208, N = 3SE +/- 0.00135, N = 3SE +/- 0.00347, N = 3SE +/- 0.00296, N = 3SE +/- 0.00749, N = 36.238008.499453.447686.056594.516504.618718.981955.88640

Java JMH

This very basic test profile runs the stock benchmark of the Java JMH benchmark via Maven. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps/s, More Is BetterJava JMHThroughputEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248816000M32000M48000M64000M80000M54316120272.1066629443540.8728953646655.5151157401824.3638412521187.6938776376503.4775134161573.0436765943472.27

Zstd Compression

This test measures the time needed to compress/decompress a sample file (silesia.tar) using Zstd (Zstandard) compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 19 - Compression SpeedEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488612182430SE +/- 0.17, N = 15SE +/- 0.48, N = 12SE +/- 0.09, N = 3SE +/- 0.23, N = 15SE +/- 0.09, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 320.022.518.024.018.318.018.816.31. (CC) gcc options: -O3 -pthread -lz -llzma

PostgreSQL

This is a benchmark of PostgreSQL using the integrated pgbench for facilitating the database benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 800 - Mode: Read Write - Average LatencyEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488714212835SE +/- 0.23, N = 3SE +/- 0.17, N = 3SE +/- 0.34, N = 3SE +/- 0.19, N = 7SE +/- 0.16, N = 3SE +/- 0.09, N = 3SE +/- 0.16, N = 3SE +/- 0.06, N = 320.0915.6831.4420.2124.7924.5016.1623.201. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 800 - Mode: Read WriteEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248811K22K33K44K55KSE +/- 454.20, N = 3SE +/- 549.43, N = 3SE +/- 280.73, N = 3SE +/- 369.77, N = 7SE +/- 206.50, N = 3SE +/- 123.75, N = 3SE +/- 481.81, N = 3SE +/- 85.58, N = 339830510222544839608322803265749522344861. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

Appleseed

Appleseed is an open-source production renderer focused on physically-based global illumination rendering engine primarily designed for animation and visual effects. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterAppleseed 2.0 BetaScene: EmilyEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248870140210280350202.53153.65341.30201.65264.67263.51152.30323.27

ScyllaDB

This is a benchmark of ScyllaDB and is making use of Apache Cassandra's cassandra-stress for conducting the benchmark. ScyllaDB is an open-source distributed NoSQL data store that is compatible with Apache Cassandra while focusing on higher throughput and lower latency. ScyllaDB uses a sharded design on each node. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterScyllaDB 5.2.9Test: WritesEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248830K60K90K120K150KSE +/- 83.55, N = 3SE +/- 369.77, N = 3SE +/- 110.37, N = 3SE +/- 415.80, N = 3SE +/- 140.30, N = 3SE +/- 60.93, N = 3SE +/- 255.51, N = 3SE +/- 395.91, N = 3112383145533780041159809334294217151729104832

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: fcn-resnet101-11 - Device: CPU - Executor: ParallelEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24882004006008001000SE +/- 7.61, N = 3SE +/- 4.73, N = 15SE +/- 8.61, N = 3SE +/- 6.77, N = 3SE +/- 10.64, N = 4SE +/- 12.91, N = 15SE +/- 1.90, N = 3SE +/- 11.74, N = 14707.63566.851042.41669.80886.83864.56567.07960.861. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: fcn-resnet101-11 - Device: CPU - Executor: ParallelEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24880.39730.79461.19191.58921.9865SE +/- 0.015352, N = 3SE +/- 0.015047, N = 15SE +/- 0.007958, N = 3SE +/- 0.014994, N = 3SE +/- 0.013735, N = 4SE +/- 0.018640, N = 15SE +/- 0.005924, N = 3SE +/- 0.013081, N = 141.4134901.7658800.9594411.4932701.1281001.1605401.7634901.0428121. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.7Preset: Very ThoroughEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24880.60391.20781.81172.41563.0195SE +/- 0.0009, N = 3SE +/- 0.0008, N = 3SE +/- 0.0004, N = 3SE +/- 0.0004, N = 3SE +/- 0.0004, N = 3SE +/- 0.0010, N = 3SE +/- 0.0093, N = 3SE +/- 0.0116, N = 31.83272.56361.03271.81641.35631.39792.68411.46391. (CXX) g++ options: -O3 -flto -pthread

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.7Preset: ExhaustiveEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24880.36880.73761.10641.47521.844SE +/- 0.0001, N = 3SE +/- 0.0003, N = 3SE +/- 0.0004, N = 3SE +/- 0.0003, N = 3SE +/- 0.0003, N = 3SE +/- 0.0000, N = 3SE +/- 0.0020, N = 3SE +/- 0.0051, N = 31.12371.57060.63301.11430.83130.85571.63910.87411. (CXX) g++ options: -O3 -flto -pthread

Apache IoTDB

Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgAverage Latency, Fewer Is BetterApache IoTDB 1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 400EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248860120180240300SE +/- 1.22, N = 3SE +/- 1.59, N = 3SE +/- 3.65, N = 3SE +/- 3.51, N = 3SE +/- 3.35, N = 3SE +/- 2.74, N = 3SE +/- 3.67, N = 3SE +/- 1.91, N = 3231.55227.34275.62231.39250.94243.61230.41275.14MAX: 27011.78MAX: 26429.57MAX: 28062.74MAX: 27094.18MAX: 27118.82MAX: 26975.13MAX: 26735.87MAX: 27185.57

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 400EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248815M30M45M60M75MSE +/- 162538.27, N = 3SE +/- 262014.17, N = 3SE +/- 374209.00, N = 3SE +/- 347434.46, N = 3SE +/- 402227.19, N = 3SE +/- 418677.10, N = 3SE +/- 489754.87, N = 3SE +/- 245819.57, N = 36877181470817373571906486943559763451680646555026973116657760197

OSPRay

Intel OSPRay is a portable ray-tracing engine for high-performance, high-fidelity scientific visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 3.1Benchmark: particle_volume/pathtracer/real_timeEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248850100150200250SE +/- 0.64, N = 3SE +/- 1.24, N = 3SE +/- 0.31, N = 3SE +/- 1.13, N = 3SE +/- 0.28, N = 3SE +/- 0.61, N = 3SE +/- 0.68, N = 3SE +/- 0.42, N = 3193.97209.22132.40182.05158.59170.39220.91157.97

Primesieve

Primesieve generates prime numbers using a highly optimized sieve of Eratosthenes implementation. Primesieve primarily benchmarks the CPU's L1/L2 cache performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterPrimesieve 12.1Length: 1e13EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248850100150200250SE +/- 0.08, N = 3SE +/- 0.07, N = 3SE +/- 0.10, N = 3SE +/- 0.05, N = 3SE +/- 0.11, N = 3SE +/- 0.15, N = 3SE +/- 0.20, N = 3SE +/- 0.18, N = 3120.8685.31205.20111.63158.45153.6082.47245.671. (CXX) g++ options: -O3

Apache IoTDB

Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgAverage Latency, Fewer Is BetterApache IoTDB 1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 100EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820406080100SE +/- 0.39, N = 3SE +/- 0.42, N = 3SE +/- 0.44, N = 3SE +/- 0.22, N = 3SE +/- 0.91, N = 3SE +/- 0.31, N = 3SE +/- 0.10, N = 3SE +/- 0.51, N = 366.4964.5279.6464.5571.6670.7065.6380.56MAX: 13829.28MAX: 12582.24MAX: 15110.81MAX: 15090.58MAX: 13850.52MAX: 13822.83MAX: 12576.01MAX: 15084.51

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 100EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248815M30M45M60M75MSE +/- 282335.19, N = 3SE +/- 248696.67, N = 3SE +/- 151013.56, N = 3SE +/- 128616.25, N = 3SE +/- 364826.83, N = 3SE +/- 42980.57, N = 3SE +/- 41672.18, N = 3SE +/- 205982.51, N = 36873598570896469579252457075257364201036647812856985751957028398

PostgreSQL

This is a benchmark of PostgreSQL using the integrated pgbench for facilitating the database benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 500 - Mode: Read Only - Average LatencyEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24880.27160.54320.81481.08641.358SE +/- 0.001, N = 3SE +/- 0.001, N = 3SE +/- 0.014, N = 3SE +/- 0.001, N = 3SE +/- 0.009, N = 3SE +/- 0.012, N = 3SE +/- 0.001, N = 3SE +/- 0.003, N = 30.5050.3601.2070.5130.9040.9610.3610.8991. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 500 - Mode: Read OnlyEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488300K600K900K1200K1500KSE +/- 2177.66, N = 3SE +/- 3546.18, N = 3SE +/- 4829.64, N = 3SE +/- 2685.11, N = 3SE +/- 5619.73, N = 3SE +/- 6448.13, N = 3SE +/- 3948.07, N = 3SE +/- 1617.01, N = 3990676138725441448597428255337152050213841665558731. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 500 - Mode: Read Write - Average LatencyEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248848121620SE +/- 0.061, N = 3SE +/- 0.089, N = 3SE +/- 0.123, N = 3SE +/- 0.080, N = 3SE +/- 0.098, N = 3SE +/- 0.078, N = 3SE +/- 0.109, N = 3SE +/- 0.101, N = 310.7148.56315.82010.90612.64412.3558.85011.7911. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 500 - Mode: Read WriteEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248813K26K39K52K65KSE +/- 263.22, N = 3SE +/- 609.40, N = 3SE +/- 246.90, N = 3SE +/- 339.91, N = 3SE +/- 304.71, N = 3SE +/- 256.18, N = 3SE +/- 699.92, N = 3SE +/- 362.37, N = 346671584053160945852395494047356517424121. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: yolov4 - Device: CPU - Executor: ParallelEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.76, N = 12SE +/- 0.77, N = 9SE +/- 1.77, N = 3SE +/- 0.13, N = 3SE +/- 1.34, N = 4SE +/- 1.14, N = 15SE +/- 1.04, N = 4SE +/- 1.78, N = 3108.9590.64147.48111.24108.48107.7986.84123.591. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: yolov4 - Device: CPU - Executor: ParallelEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24883691215SE +/- 0.06480, N = 12SE +/- 0.09061, N = 9SE +/- 0.08117, N = 3SE +/- 0.01052, N = 3SE +/- 0.11284, N = 4SE +/- 0.10193, N = 15SE +/- 0.13812, N = 4SE +/- 0.11600, N = 39.1834011.038006.782268.989589.222199.2917911.519908.094211. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: SP.CEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24883K6K9K12K15KSE +/- 27.04, N = 3SE +/- 188.19, N = 3SE +/- 7.67, N = 3SE +/- 10.77, N = 3SE +/- 5.49, N = 3SE +/- 1.44, N = 3SE +/- 146.76, N = 5SE +/- 4.20, N = 311606.2215940.6110028.7414130.3010088.5010027.1113374.3511387.051. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6

FFmpeg

This is a benchmark of the FFmpeg multimedia framework. The FFmpeg test profile is making use of a modified version of vbench from Columbia University's Architecture and Design Lab (ARCADE) [http://arcade.cs.columbia.edu/vbench/] that is a benchmark for video-as-a-service workloads. The test profile offers the options of a range of vbench scenarios based on freely distributable video content and offers the options of using the x264 or x265 video encoders for transcoding. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 6.1Encoder: libx265 - Scenario: PlatformEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881632486480SE +/- 0.03, N = 3SE +/- 0.19, N = 3SE +/- 0.01, N = 3SE +/- 0.06, N = 3SE +/- 0.08, N = 3SE +/- 0.05, N = 3SE +/- 0.10, N = 3SE +/- 0.22, N = 365.8265.4960.2363.6864.7765.2669.8058.751. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 6.1Encoder: libx265 - Scenario: Video On DemandEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881632486480SE +/- 0.04, N = 3SE +/- 0.10, N = 3SE +/- 0.08, N = 3SE +/- 0.09, N = 3SE +/- 0.05, N = 3SE +/- 0.09, N = 3SE +/- 0.13, N = 3SE +/- 0.02, N = 365.8765.2460.2063.7064.9265.3569.8158.751. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 6.1Encoder: libx265 - Scenario: UploadEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488816243240SE +/- 0.05, N = 3SE +/- 0.06, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.07, N = 3SE +/- 0.08, N = 332.4032.1429.1731.3231.7331.9534.2128.891. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: LU.CEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248810K20K30K40K50KSE +/- 221.05, N = 3SE +/- 266.87, N = 3SE +/- 11.58, N = 3SE +/- 13.15, N = 3SE +/- 38.55, N = 3SE +/- 613.23, N = 15SE +/- 521.71, N = 15SE +/- 714.33, N = 1541255.9848014.8626620.6943166.1040284.2337149.8847556.6940491.561. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6

DuckDB

DuckDB is an in-progress SQL OLAP database management system optimized for analytics and features a vectorized and parallel engine. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterDuckDB 0.9.1Benchmark: IMDBEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.29, N = 3SE +/- 0.07, N = 3SE +/- 0.05, N = 3SE +/- 0.05, N = 3SE +/- 0.07, N = 3SE +/- 0.06, N = 3SE +/- 0.05, N = 3SE +/- 0.14, N = 390.0383.72115.6487.84101.97100.6382.3899.581. (CXX) g++ options: -O3 -rdynamic -lssl -lcrypto -ldl

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: T5 Encoder - Device: CPU - Executor: StandardEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488246810SE +/- 0.07202, N = 4SE +/- 0.03228, N = 3SE +/- 0.07773, N = 4SE +/- 0.05281, N = 15SE +/- 0.05571, N = 3SE +/- 0.03666, N = 3SE +/- 0.08220, N = 3SE +/- 0.07186, N = 155.932075.018246.572604.900576.967427.167516.161455.907861. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: T5 Encoder - Device: CPU - Executor: StandardEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24884080120160200SE +/- 2.03, N = 4SE +/- 1.28, N = 3SE +/- 1.77, N = 4SE +/- 2.12, N = 15SE +/- 1.14, N = 3SE +/- 0.71, N = 3SE +/- 2.20, N = 3SE +/- 2.25, N = 15168.62199.25152.18204.34143.50139.48162.33169.621. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

LuxCoreRender

LuxCoreRender is an open-source 3D physically based renderer formerly known as LuxRender. LuxCoreRender supports CPU-based rendering as well as GPU acceleration via OpenCL, NVIDIA CUDA, and NVIDIA OptiX interfaces. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: Danish Mood - Acceleration: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24880.94051.8812.82153.7624.7025SE +/- 0.03, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 5SE +/- 0.03, N = 5SE +/- 0.02, N = 3SE +/- 0.02, N = 11SE +/- 0.04, N = 3SE +/- 0.01, N = 152.774.151.462.812.082.104.181.45MIN: 0.99 / MAX: 3.25MIN: 1.74 / MAX: 4.72MIN: 0.38 / MAX: 1.85MIN: 1.01 / MAX: 3.32MIN: 0.66 / MAX: 2.47MIN: 0.64 / MAX: 2.54MIN: 1.81 / MAX: 4.75MIN: 0.39 / MAX: 1.82

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Junkshop - Compute: CPU-OnlyEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24884080120160200SE +/- 0.08, N = 3SE +/- 0.10, N = 3SE +/- 0.25, N = 3SE +/- 0.15, N = 3SE +/- 0.48, N = 3SE +/- 0.43, N = 3SE +/- 0.17, N = 3SE +/- 0.02, N = 3103.2572.78186.24101.19141.77139.4872.75173.17

SPECFEM3D

simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra. This test profile currently relies on CPU-based execution for SPECFEM3D and using a variety of their built-in examples/models for benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.1.1Model: Water-layered HalfspaceEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.52, N = 3SE +/- 0.54, N = 3SE +/- 0.07, N = 3SE +/- 0.24, N = 3SE +/- 0.81, N = 3SE +/- 0.42, N = 3SE +/- 0.06, N = 3SE +/- 0.21, N = 3104.1278.38156.91131.87130.20129.3385.25135.421. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.1.1Model: Layered HalfspaceEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24884080120160200SE +/- 0.17, N = 3SE +/- 0.13, N = 3SE +/- 1.62, N = 3SE +/- 0.40, N = 3SE +/- 1.33, N = 3SE +/- 0.49, N = 3SE +/- 0.14, N = 3SE +/- 1.26, N = 3103.7381.88166.97113.17133.47133.3183.97144.541. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

OSPRay

Intel OSPRay is a portable ray-tracing engine for high-performance, high-fidelity scientific visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 3.1Benchmark: particle_volume/ao/real_timeEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24883691215SE +/- 0.00160, N = 3SE +/- 0.00180, N = 3SE +/- 0.00096, N = 3SE +/- 0.00892, N = 3SE +/- 0.00247, N = 3SE +/- 0.00129, N = 3SE +/- 0.00600, N = 3SE +/- 0.00839, N = 36.254188.498693.455776.068954.525524.638288.983995.93161

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: GPT-2 - Device: CPU - Executor: StandardEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488246810SE +/- 0.07799, N = 4SE +/- 0.01007, N = 3SE +/- 0.00493, N = 3SE +/- 0.13250, N = 12SE +/- 0.08771, N = 4SE +/- 0.01376, N = 3SE +/- 0.08068, N = 15SE +/- 0.02440, N = 37.092407.121007.434686.205257.421268.063897.517447.890801. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: GPT-2 - Device: CPU - Executor: StandardEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24884080120160200SE +/- 1.56, N = 4SE +/- 0.20, N = 3SE +/- 0.08, N = 3SE +/- 3.26, N = 12SE +/- 1.56, N = 4SE +/- 0.21, N = 3SE +/- 1.39, N = 15SE +/- 0.39, N = 3140.98140.35134.45161.84134.74123.95133.17126.671. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

Intel Open Image Denoise

Open Image Denoise is a denoising library for ray-tracing and part of the Intel oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 2.2Run: RTLightmap.hdr.4096x4096 - Device: CPU-OnlyEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24880.09230.18460.27690.36920.4615SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.300.400.170.300.220.230.410.21

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Fishy Cat - Compute: CPU-OnlyEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24884080120160200SE +/- 0.16, N = 3SE +/- 0.12, N = 3SE +/- 0.15, N = 3SE +/- 0.02, N = 3SE +/- 0.16, N = 3SE +/- 0.39, N = 3SE +/- 0.06, N = 3SE +/- 0.24, N = 397.8467.99170.0095.44130.42127.0166.78165.00

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 32 - Model: ResNet-50EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488918273645SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 334.7240.3725.7034.7831.0030.7939.9923.19

NAMD

NAMD is a parallel molecular dynamics code designed for high-performance simulation of large biomolecular systems. NAMD was developed by the Theoretical and Computational Biophysics Group in the Beckman Institute for Advanced Science and Technology at the University of Illinois at Urbana-Champaign. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgns/day, More Is BetterNAMD 3.0b6Input: STMV with 1,066,628 AtomsEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24880.17050.3410.51150.6820.8525SE +/- 0.00106, N = 3SE +/- 0.00443, N = 3SE +/- 0.00046, N = 3SE +/- 0.00185, N = 3SE +/- 0.00019, N = 3SE +/- 0.00059, N = 3SE +/- 0.00230, N = 3SE +/- 0.00016, N = 30.608130.757680.389560.270390.485700.493450.756460.20565

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 256 - Model: AlexNetEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248890180270360450SE +/- 0.07, N = 3SE +/- 0.23, N = 3SE +/- 0.14, N = 3SE +/- 0.35, N = 3SE +/- 0.05, N = 3SE +/- 0.09, N = 3SE +/- 0.09, N = 3SE +/- 0.06, N = 3305.22396.85174.06294.91235.47237.57409.96189.14

libavif avifenc

This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 1.0Encoder Speed: 0EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.10, N = 3SE +/- 0.77, N = 5SE +/- 0.29, N = 3SE +/- 1.17, N = 3SE +/- 0.27, N = 3SE +/- 0.18, N = 3SE +/- 0.33, N = 3SE +/- 0.49, N = 389.6374.06136.6391.59110.46108.6572.11133.571. (CXX) g++ options: -O3 -fPIC -lm

CockroachDB

CockroachDB is a cloud-native, distributed SQL database for data intensive applications. This test profile uses a server-less CockroachDB configuration to test various Coackroach workloads on the local host with a single node. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgops/s, More Is BetterCockroachDB 22.2Workload: KV, 50% Reads - Concurrency: 128EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820K40K60K80K100KSE +/- 37.01, N = 3SE +/- 58.30, N = 3SE +/- 67.60, N = 3SE +/- 233.77, N = 3SE +/- 75.55, N = 3SE +/- 264.18, N = 3SE +/- 831.25, N = 3SE +/- 189.35, N = 371721.190314.049645.069839.863986.264052.792195.259866.6

OpenBenchmarking.orgops/s, More Is BetterCockroachDB 22.2Workload: KV, 60% Reads - Concurrency: 128EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820K40K60K80K100KSE +/- 132.40, N = 3SE +/- 462.84, N = 3SE +/- 116.97, N = 3SE +/- 72.81, N = 3SE +/- 105.77, N = 3SE +/- 47.09, N = 3SE +/- 156.58, N = 3SE +/- 198.29, N = 375445.796269.151927.073004.566756.967506.795084.863021.4

OpenBenchmarking.orgops/s, More Is BetterCockroachDB 22.2Workload: KV, 95% Reads - Concurrency: 128EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248830K60K90K120K150KSE +/- 58.24, N = 3SE +/- 158.53, N = 3SE +/- 73.92, N = 3SE +/- 391.29, N = 3SE +/- 276.11, N = 3SE +/- 302.28, N = 3SE +/- 765.52, N = 3SE +/- 237.27, N = 392280.7118066.763936.290283.682706.283155.8121061.783987.5

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: bertsquad-12 - Device: CPU - Executor: ParallelEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820406080100SE +/- 0.64, N = 9SE +/- 0.46, N = 14SE +/- 0.30, N = 3SE +/- 0.95, N = 3SE +/- 0.40, N = 3SE +/- 0.76, N = 3SE +/- 0.22, N = 3SE +/- 0.78, N = 378.3967.76109.2479.3181.1579.7064.2886.691. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: bertsquad-12 - Device: CPU - Executor: ParallelEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248848121620SE +/- 0.10273, N = 9SE +/- 0.09843, N = 14SE +/- 0.02536, N = 3SE +/- 0.15346, N = 3SE +/- 0.06035, N = 3SE +/- 0.11789, N = 3SE +/- 0.05371, N = 3SE +/- 0.10358, N = 312.7624014.766409.1545812.6118012.3239012.5483015.5560011.537201. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

VVenC

VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVVenC 1.11Video Input: Bosphorus 4K - Video Preset: FastEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488246810SE +/- 0.029, N = 3SE +/- 0.045, N = 3SE +/- 0.017, N = 3SE +/- 0.034, N = 3SE +/- 0.031, N = 3SE +/- 0.021, N = 3SE +/- 0.058, N = 3SE +/- 0.018, N = 36.4907.4944.5426.4525.5545.6937.8184.9751. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects

GROMACS

The GROMACS (GROningen MAchine for Chemical Simulations) molecular dynamics package testing with the water_GMX50 data. This test profile allows selecting between CPU and GPU-based GROMACS builds. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNs Per Day, More Is BetterGROMACS 2024Implementation: MPI CPU - Input: water_GMX50_bareEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24880.60351.2071.81052.4143.0175SE +/- 0.008, N = 3SE +/- 0.002, N = 3SE +/- 0.001, N = 3SE +/- 0.001, N = 3SE +/- 0.002, N = 3SE +/- 0.002, N = 3SE +/- 0.001, N = 3SE +/- 0.000, N = 32.1232.6821.2532.2881.5221.5462.5431.4341. (CXX) g++ options: -O3 -lm

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: BT.CEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248810K20K30K40K50KSE +/- 229.09, N = 3SE +/- 481.60, N = 5SE +/- 5.49, N = 3SE +/- 193.15, N = 3SE +/- 20.15, N = 3SE +/- 14.25, N = 3SE +/- 110.88, N = 3SE +/- 96.46, N = 337547.7647856.8722331.4640212.6522976.5322935.2744844.1328326.361. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6

Numpy Benchmark

This is a test to obtain the general Numpy performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgScore, More Is BetterNumpy BenchmarkEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24882004006008001000SE +/- 0.83, N = 3SE +/- 3.03, N = 3SE +/- 2.19, N = 3SE +/- 0.68, N = 3SE +/- 1.60, N = 3SE +/- 3.59, N = 3SE +/- 2.05, N = 3SE +/- 1.43, N = 3812.29746.24773.69791.79804.51808.62803.04675.48

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 3 - Resolution: 1080p - Samples Per Pixel: 16 - Renderer: Path Tracer - Acceleration: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248813K26K39K52K65KSE +/- 102.05, N = 3SE +/- 45.16, N = 3SE +/- 91.50, N = 3SE +/- 529.04, N = 15SE +/- 85.35, N = 3SE +/- 62.68, N = 3SE +/- 29.87, N = 3SE +/- 115.84, N = 33296821517584423111644674438302028348929

ACES DGEMM

This is a multi-threaded DGEMM benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterACES DGEMM 1.0Sustained Floating-Point RateEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24883691215SE +/- 0.115135, N = 15SE +/- 0.322380, N = 15SE +/- 0.039171, N = 3SE +/- 0.066520, N = 3SE +/- 0.037881, N = 3SE +/- 0.144416, N = 15SE +/- 0.258971, N = 12SE +/- 0.054700, N = 37.18706210.5916513.9201685.6045025.3136495.3305009.5777534.2041101. (CC) gcc options: -O3 -march=native -fopenmp

PyTorch

This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 64 - Model: ResNet-50EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881020304050SE +/- 0.17, N = 3SE +/- 0.34, N = 3SE +/- 0.17, N = 3SE +/- 0.01, N = 3SE +/- 0.14, N = 3SE +/- 0.16, N = 3SE +/- 0.94, N = 15SE +/- 1.58, N = 1539.5346.2034.8841.3443.8945.1143.3832.25MIN: 38.01 / MAX: 40.46MIN: 43.63 / MAX: 47.16MIN: 21.67 / MAX: 35.54MIN: 30.45 / MAX: 41.8MIN: 42.37 / MAX: 45.05MIN: 43.61 / MAX: 45.86MIN: 34.95 / MAX: 46.71MIN: 23.14 / MAX: 40.93

Zstd Compression

This test measures the time needed to compress/decompress a sample file (silesia.tar) using Zstd (Zstandard) compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 12 - Compression SpeedEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248860120180240300SE +/- 2.73, N = 3SE +/- 6.73, N = 15SE +/- 0.43, N = 3SE +/- 1.66, N = 3SE +/- 0.62, N = 3SE +/- 0.92, N = 3SE +/- 2.97, N = 3SE +/- 2.27, N = 3221.2248.0164.4273.1177.9171.0220.9170.61. (CC) gcc options: -O3 -pthread -lz -llzma

Appleseed

Appleseed is an open-source production renderer focused on physically-based global illumination rendering engine primarily designed for animation and visual effects. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterAppleseed 2.0 BetaScene: Material TesterEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24884080120160200129.09104.32191.21131.17146.67142.94102.15173.41

OpenBenchmarking.orgSeconds, Fewer Is BetterAppleseed 2.0 BetaScene: Disney MaterialEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248850100150200250120.9187.70210.11118.62160.01156.0785.32174.95

ASKAP

ASKAP is a set of benchmarks from the Australian SKA Pathfinder. The principal ASKAP benchmarks are the Hogbom Clean Benchmark (tHogbomClean) and Convolutional Resamping Benchmark (tConvolve) as well as some previous ASKAP benchmarks being included as well for OpenCL and CUDA execution of tConvolve. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMpix/sec, More Is BetterASKAP 1.0Test: tConvolve MPI - GriddingEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24882K4K6K8K10KSE +/- 528.48, N = 12SE +/- 384.12, N = 15SE +/- 0.00, N = 3SE +/- 175.34, N = 15SE +/- 44.18, N = 3SE +/- 180.65, N = 12SE +/- 298.02, N = 15SE +/- 48.43, N = 37389.249760.755963.547529.935875.194256.287477.196199.091. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp

OpenBenchmarking.orgMpix/sec, More Is BetterASKAP 1.0Test: tConvolve MPI - DegriddingEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24882K4K6K8K10KSE +/- 473.28, N = 12SE +/- 346.60, N = 15SE +/- 38.09, N = 3SE +/- 187.79, N = 15SE +/- 34.30, N = 3SE +/- 192.19, N = 12SE +/- 321.53, N = 15SE +/- 58.27, N = 36270.828364.074723.736140.895179.313635.696642.945146.331. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp

MariaDB mariadb-slap

This is a MariaDB MySQL database server benchmark making use of mysqlslap (mariadb-slap). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB mariadb-slap 11.5Clients: 64EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248830060090012001500SE +/- 7.01, N = 3SE +/- 4.14, N = 3SE +/- 3.77, N = 3SE +/- 5.80, N = 3SE +/- 3.26, N = 3SE +/- 5.66, N = 3SE +/- 4.85, N = 3SE +/- 4.64, N = 311311124100910761095112411831128-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto1. (CXX) g++ options: -fPIC -pie -fstack-protector -O3 -lpthread -lz -ldl -lm

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: BMW27 - Compute: CPU-OnlyEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.16, N = 3SE +/- 0.14, N = 3SE +/- 0.46, N = 3SE +/- 0.17, N = 3SE +/- 0.10, N = 3SE +/- 0.33, N = 3SE +/- 0.05, N = 3SE +/- 0.18, N = 378.0954.64136.8376.39104.74101.7253.25123.51

nginx

This is a benchmark of the lightweight Nginx HTTP(S) web-server. This Nginx web server benchmark test profile makes use of the wrk program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients/connections. HTTPS with a self-signed OpenSSL certificate is used by this test for local benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.23.2Connections: 1000EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248830K60K90K120K150KSE +/- 117.22, N = 3SE +/- 280.22, N = 3SE +/- 150.02, N = 3SE +/- 194.32, N = 3SE +/- 87.10, N = 3SE +/- 98.21, N = 3SE +/- 301.50, N = 3SE +/- 46.60, N = 3108719.12150874.9166910.93120690.4384832.8285990.12136360.7287132.221. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.23.2Connections: 500EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248830K60K90K120K150KSE +/- 41.74, N = 3SE +/- 336.44, N = 3SE +/- 14.73, N = 3SE +/- 88.44, N = 3SE +/- 64.09, N = 3SE +/- 191.32, N = 3SE +/- 288.60, N = 3SE +/- 173.47, N = 3116236.16158717.7470866.42124133.8891569.8794947.75154835.0692582.291. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.23.2Connections: 200EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248830K60K90K120K150KSE +/- 69.09, N = 3SE +/- 28.66, N = 3SE +/- 50.61, N = 3SE +/- 32.54, N = 3SE +/- 78.04, N = 3SE +/- 127.61, N = 3SE +/- 150.85, N = 3SE +/- 231.66, N = 3118177.50160748.1875062.51123864.1697227.42101214.88162348.3398692.951. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2

OpenFOAM

OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Small Mesh Size - Execution TimeEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248870140210280350210.39141.46307.01168.98275.10296.60183.23242.941. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfoamToVTK -ldynamicMesh -llagrangian -lgenericPatchFields -lfileFormats -lOpenFOAM -ldl -lm

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Small Mesh Size - Mesh TimeEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248891827364532.8529.6939.1633.6533.3338.1730.7735.191. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfoamToVTK -ldynamicMesh -llagrangian -lgenericPatchFields -lfileFormats -lOpenFOAM -ldl -lm

Timed Linux Kernel Compilation

This test times how long it takes to build the Linux kernel in a default configuration (defconfig) for the architecture being tested or alternatively an allmodconfig for building all possible kernel modules for the build. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.8Build: defconfigEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.47, N = 3SE +/- 0.35, N = 3SE +/- 0.29, N = 3SE +/- 0.38, N = 3SE +/- 0.40, N = 3SE +/- 0.32, N = 3SE +/- 0.36, N = 3SE +/- 0.45, N = 375.2557.48126.0070.3098.2997.4157.47108.34

Pennant

Pennant is an application focused on hydrodynamics on general unstructured meshes in 2D. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgHydro Cycle Time - Seconds, Fewer Is BetterPennant 1.0.1Test: leblancbigEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881020304050SE +/- 0.42, N = 15SE +/- 0.17, N = 3SE +/- 0.11, N = 3SE +/- 0.35, N = 3SE +/- 0.00, N = 3SE +/- 0.88, N = 15SE +/- 0.27, N = 3SE +/- 1.09, N = 1231.8622.9741.7128.1435.4842.9429.5532.871. (CXX) g++ options: -fopenmp -lmpi_cxx -lmpi

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 3 - Resolution: 1080p - Samples Per Pixel: 32 - Renderer: Path Tracer - Acceleration: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820K40K60K80K100KSE +/- 30.37, N = 3SE +/- 195.52, N = 3SE +/- 16.70, N = 3SE +/- 84.90, N = 3SE +/- 185.23, N = 3SE +/- 273.73, N = 3SE +/- 101.69, N = 3SE +/- 76.92, N = 361937469041135016468985743842094424795447

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: Recurrent Neural Network Training - Engine: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24887001400210028003500SE +/- 2.12, N = 3SE +/- 1.75, N = 3SE +/- 0.60, N = 3SE +/- 1.07, N = 3SE +/- 3.47, N = 3SE +/- 1.64, N = 3SE +/- 1.43, N = 3SE +/- 9.27, N = 31973.991355.103148.811890.002514.662491.141348.372931.95MIN: 1936.17MIN: 1323.99MIN: 3143.3MIN: 1885.76MIN: 2499.44MIN: 2459.9MIN: 1314.05MIN: 2882.151. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

CloverLeaf

CloverLeaf is a Lagrangian-Eulerian hydrodynamics benchmark. This test profile currently makes use of CloverLeaf's OpenMP version. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterCloverLeaf 1.3Input: clover_bmEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820406080100SE +/- 0.28, N = 3SE +/- 0.47, N = 6SE +/- 0.13, N = 3SE +/- 0.10, N = 3SE +/- 0.05, N = 3SE +/- 0.21, N = 3SE +/- 0.45, N = 3SE +/- 0.27, N = 369.9847.0277.2385.7374.5986.8272.8864.581. (F9X) gfortran options: -O3 -march=native -funroll-loops -fopenmp

Zstd Compression

This test measures the time needed to compress/decompress a sample file (silesia.tar) using Zstd (Zstandard) compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 8 - Compression SpeedEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24882004006008001000SE +/- 2.23, N = 3SE +/- 2.98, N = 3SE +/- 2.29, N = 3SE +/- 6.53, N = 3SE +/- 1.07, N = 3SE +/- 1.88, N = 3SE +/- 6.27, N = 9SE +/- 3.15, N = 3798.4970.7500.0942.3525.1520.3793.8485.61. (CC) gcc options: -O3 -pthread -lz -llzma

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: CaffeNet 12-int8 - Device: CPU - Executor: ParallelEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24880.40350.8071.21051.6142.0175SE +/- 0.01429, N = 3SE +/- 0.01034, N = 7SE +/- 0.01689, N = 6SE +/- 0.00881, N = 3SE +/- 0.00514, N = 3SE +/- 0.01555, N = 3SE +/- 0.01197, N = 3SE +/- 0.01234, N = 31.387921.115521.712571.298191.486781.475311.205041.793471. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: CaffeNet 12-int8 - Device: CPU - Executor: ParallelEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24882004006008001000SE +/- 7.33, N = 3SE +/- 8.12, N = 7SE +/- 5.55, N = 6SE +/- 5.20, N = 3SE +/- 2.32, N = 3SE +/- 7.12, N = 3SE +/- 8.14, N = 3SE +/- 3.86, N = 3719.99895.93583.98769.69672.35677.58829.19557.411. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

Zstd Compression

This test measures the time needed to compress/decompress a sample file (silesia.tar) using Zstd (Zstandard) compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 19, Long Mode - Compression SpeedEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24883691215SE +/- 0.15, N = 3SE +/- 0.14, N = 5SE +/- 0.00, N = 3SE +/- 0.15, N = 3SE +/- 0.06, N = 3SE +/- 0.00, N = 3SE +/- 0.09, N = 3SE +/- 0.01, N = 311.5012.9010.2013.2010.5010.2010.708.911. (CC) gcc options: -O3 -pthread -lz -llzma

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: Recurrent Neural Network Inference - Engine: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248830060090012001500SE +/- 1.03, N = 3SE +/- 2.47, N = 3SE +/- 0.51, N = 3SE +/- 0.47, N = 3SE +/- 0.53, N = 3SE +/- 7.18, N = 3SE +/- 1.91, N = 3SE +/- 6.38, N = 31036.15694.521615.38962.211310.591337.47706.651548.71MIN: 1006.91MIN: 671.73MIN: 1610.14MIN: 959.35MIN: 1303.57MIN: 1299.81MIN: 679.48MIN: 1512.561. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 1 - Buffer Length: 256 - Filter Length: 512EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24884M8M12M16M20MSE +/- 421403.42, N = 15SE +/- 267438.19, N = 15SE +/- 150617.99, N = 3SE +/- 44508.43, N = 3SE +/- 189549.47, N = 3SE +/- 250051.55, N = 3SE +/- 410224.76, N = 15SE +/- 13650.40, N = 318078200191190671842633319125000193650001961166718964400202550001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

NAMD

NAMD is a parallel molecular dynamics code designed for high-performance simulation of large biomolecular systems. NAMD was developed by the Theoretical and Computational Biophysics Group in the Beckman Institute for Advanced Science and Technology at the University of Illinois at Urbana-Champaign. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgns/day, More Is BetterNAMD 3.0b6Input: ATPase with 327,506 AtomsEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24880.65521.31041.96562.62083.276SE +/- 0.00983, N = 3SE +/- 0.00837, N = 3SE +/- 0.00340, N = 3SE +/- 0.01721, N = 15SE +/- 0.00076, N = 3SE +/- 0.00212, N = 3SE +/- 0.03311, N = 4SE +/- 0.00164, N = 32.205602.911831.319210.654251.691201.737642.788460.65977

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 3 - Resolution: 1080p - Samples Per Pixel: 1 - Renderer: Path Tracer - Acceleration: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24887001400210028003500SE +/- 3.18, N = 3SE +/- 15.53, N = 4SE +/- 5.49, N = 3SE +/- 2.33, N = 3SE +/- 2.52, N = 3SE +/- 4.04, N = 3SE +/- 0.33, N = 3SE +/- 0.88, N = 318311319343018982575252612652884

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 64 - Model: GoogLeNetEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.08, N = 3SE +/- 0.04, N = 3SE +/- 0.01, N = 3SE +/- 0.08, N = 3SE +/- 0.05, N = 3SE +/- 0.04, N = 3SE +/- 0.07, N = 3SE +/- 0.03, N = 3110.52135.7873.77113.5191.2192.37133.2178.17

SPECFEM3D

simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra. This test profile currently relies on CPU-based execution for SPECFEM3D and using a variety of their built-in examples/models for benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.1.1Model: Mount St. HelensEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881428425670SE +/- 0.26, N = 3SE +/- 0.35, N = 5SE +/- 0.61, N = 5SE +/- 0.12, N = 3SE +/- 0.47, N = 3SE +/- 0.25, N = 3SE +/- 0.29, N = 15SE +/- 0.34, N = 339.0131.1262.5747.8551.5951.2933.1353.291. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

Chaos Group V-RAY

This is a test of Chaos Group's V-RAY benchmark. V-RAY is a commercial renderer that can integrate with various creator software products like SketchUp and 3ds Max. The V-RAY benchmark is standalone and supports CPU and NVIDIA CUDA/RTX based rendering. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgvsamples, More Is BetterChaos Group V-RAY 6.0Mode: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24889K18K27K36K45KSE +/- 61.57, N = 3SE +/- 96.96, N = 3SE +/- 50.00, N = 3SE +/- 54.69, N = 3SE +/- 22.00, N = 3SE +/- 11.00, N = 3SE +/- 204.38, N = 3SE +/- 16.84, N = 33088543767172733139922505230784387920060

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 2 - Resolution: 1080p - Samples Per Pixel: 32 - Renderer: Path Tracer - Acceleration: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820K40K60K80K100KSE +/- 126.62, N = 3SE +/- 140.94, N = 3SE +/- 222.81, N = 3SE +/- 95.50, N = 3SE +/- 84.09, N = 3SE +/- 152.10, N = 3SE +/- 61.02, N = 3SE +/- 34.91, N = 35372639089973075420273858722503819282403

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: ArcFace ResNet-100 - Device: CPU - Executor: ParallelEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881122334455SE +/- 0.28, N = 7SE +/- 0.10, N = 3SE +/- 0.30, N = 3SE +/- 0.10, N = 3SE +/- 0.25, N = 3SE +/- 0.15, N = 3SE +/- 0.04, N = 3SE +/- 0.21, N = 333.1428.8050.4134.0537.2337.0026.8343.401. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: ArcFace ResNet-100 - Device: CPU - Executor: ParallelEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488918273645SE +/- 0.27, N = 7SE +/- 0.12, N = 3SE +/- 0.12, N = 3SE +/- 0.08, N = 3SE +/- 0.18, N = 3SE +/- 0.11, N = 3SE +/- 0.05, N = 3SE +/- 0.11, N = 330.1934.7219.8429.3726.8627.0237.2723.041. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 1 - Resolution: 1080p - Samples Per Pixel: 32 - Renderer: Path Tracer - Acceleration: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820K40K60K80K100KSE +/- 35.09, N = 3SE +/- 55.84, N = 3SE +/- 217.15, N = 3SE +/- 40.93, N = 3SE +/- 106.16, N = 3SE +/- 145.56, N = 3SE +/- 168.15, N = 3SE +/- 135.20, N = 35287240142960235342672930715303792381236

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 1 - Resolution: 1080p - Samples Per Pixel: 1 - Renderer: Path Tracer - Acceleration: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24886001200180024003000SE +/- 1.45, N = 3SE +/- 1.53, N = 3SE +/- 1.45, N = 3SE +/- 16.15, N = 3SE +/- 1.86, N = 3SE +/- 1.76, N = 3SE +/- 1.00, N = 3SE +/- 0.88, N = 315541140289915902176213310692441

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 2 - Resolution: 1080p - Samples Per Pixel: 1 - Renderer: Path Tracer - Acceleration: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24886001200180024003000SE +/- 0.58, N = 3SE +/- 0.58, N = 3SE +/- 3.28, N = 3SE +/- 9.29, N = 3SE +/- 0.67, N = 3SE +/- 0.33, N = 3SE +/- 2.08, N = 3SE +/- 2.33, N = 315621155293015892201215610802474

Pennant

Pennant is an application focused on hydrodynamics on general unstructured meshes in 2D. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgHydro Cycle Time - Seconds, Fewer Is BetterPennant 1.0.1Test: sedovbigEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881428425670SE +/- 0.18, N = 3SE +/- 0.12, N = 3SE +/- 0.22, N = 3SE +/- 0.43, N = 3SE +/- 0.05, N = 3SE +/- 0.59, N = 6SE +/- 0.35, N = 8SE +/- 0.67, N = 449.1834.6459.0242.1850.8961.9742.2355.901. (CXX) g++ options: -fopenmp -lmpi_cxx -lmpi

OSPRay

Intel OSPRay is a portable ray-tracing engine for high-performance, high-fidelity scientific visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 3.1Benchmark: gravity_spheres_volume/dim_512/pathtracer/real_timeEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24883691215SE +/- 0.00517, N = 3SE +/- 0.00316, N = 3SE +/- 0.00315, N = 3SE +/- 0.00278, N = 3SE +/- 0.00576, N = 3SE +/- 0.00236, N = 3SE +/- 0.00353, N = 3SE +/- 0.00822, N = 36.316058.668323.496996.091484.572224.667578.978133.79539

Timed CPython Compilation

This test times how long it takes to build the reference Python implementation, CPython, with optimizations and LTO enabled for a release build. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed CPython Compilation 3.10.6Build Configuration: Released Build, PGO + LTO OptimizedEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248850100150200250197.98198.06214.38199.92202.26200.98193.47197.92

Memcached

Memcached is a high performance, distributed memory object caching system. This Memcached test profiles makes use of memtier_benchmark for excuting this CPU/memory-focused server benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps/sec, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:10EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488800K1600K2400K3200K4000KSE +/- 1507.35, N = 3SE +/- 5211.99, N = 3SE +/- 1442.09, N = 3SE +/- 4264.77, N = 3SE +/- 2995.56, N = 3SE +/- 2761.61, N = 3SE +/- 5303.54, N = 3SE +/- 7388.22, N = 32762180.283693215.301651621.672658380.052163987.012195688.493761626.613072373.871. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

OpenBenchmarking.orgOps/sec, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:100EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488800K1600K2400K3200K4000KSE +/- 2341.26, N = 3SE +/- 8167.06, N = 3SE +/- 2895.46, N = 3SE +/- 4210.88, N = 3SE +/- 390.53, N = 3SE +/- 1471.26, N = 3SE +/- 5665.65, N = 3SE +/- 11523.53, N = 32749241.803756154.981567216.542651410.812056776.092104490.553844031.283048853.561. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: MediumEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248848121620SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 310.4115.076.5310.258.358.6115.926.62

OSPRay

Intel OSPRay is a portable ray-tracing engine for high-performance, high-fidelity scientific visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 3.1Benchmark: gravity_spheres_volume/dim_512/ao/real_timeEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488246810SE +/- 0.00993, N = 3SE +/- 0.03021, N = 3SE +/- 0.00463, N = 3SE +/- 0.01719, N = 3SE +/- 0.00148, N = 3SE +/- 0.00682, N = 3SE +/- 0.00259, N = 3SE +/- 0.00479, N = 35.347537.307482.950105.185703.816913.910707.466492.73496

OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 3.1Benchmark: gravity_spheres_volume/dim_512/scivis/real_timeEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488246810SE +/- 0.01288, N = 3SE +/- 0.00808, N = 3SE +/- 0.01153, N = 3SE +/- 0.01525, N = 3SE +/- 0.01618, N = 3SE +/- 0.00260, N = 3SE +/- 0.02548, N = 3SE +/- 0.00622, N = 35.273337.144702.919515.105893.794883.821387.369092.63851

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: Faster R-CNN R-50-FPN-int8 - Device: CPU - Executor: ParallelEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488612182430SE +/- 0.19, N = 3SE +/- 0.20, N = 4SE +/- 0.07, N = 3SE +/- 0.03, N = 3SE +/- 0.13, N = 3SE +/- 0.17, N = 3SE +/- 0.07, N = 3SE +/- 0.18, N = 320.1818.8221.4720.7720.0819.7918.9724.391. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: Faster R-CNN R-50-FPN-int8 - Device: CPU - Executor: ParallelEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881224364860SE +/- 0.46, N = 3SE +/- 0.57, N = 4SE +/- 0.15, N = 3SE +/- 0.08, N = 3SE +/- 0.32, N = 3SE +/- 0.43, N = 3SE +/- 0.21, N = 3SE +/- 0.30, N = 349.5553.1546.5648.1449.8150.5352.7241.011. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.7Preset: ThoroughEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488510152025SE +/- 0.0021, N = 3SE +/- 0.0075, N = 3SE +/- 0.0018, N = 3SE +/- 0.0127, N = 3SE +/- 0.0050, N = 3SE +/- 0.0128, N = 3SE +/- 0.0627, N = 3SE +/- 0.3204, N = 1213.295318.61757.513613.17249.867410.208919.903812.23071. (CXX) g++ options: -O3 -flto -pthread

LuxCoreRender

LuxCoreRender is an open-source 3D physically based renderer formerly known as LuxRender. LuxCoreRender supports CPU-based rendering as well as GPU acceleration via OpenCL, NVIDIA CUDA, and NVIDIA OptiX interfaces. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: Orange Juice - Acceleration: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488246810SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 35.557.653.305.554.304.427.723.58MIN: 5.2 / MAX: 5.66MIN: 6.89 / MAX: 7.99MIN: 3.26 / MAX: 3.33MIN: 5.19 / MAX: 5.69MIN: 4.24 / MAX: 4.36MIN: 4.32 / MAX: 4.49MIN: 6.87 / MAX: 8.15MIN: 3.44 / MAX: 4.42

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: ResNet50 v1-12-int8 - Device: CPU - Executor: ParallelEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881.04862.09723.14584.19445.243SE +/- 0.00255, N = 3SE +/- 0.00924, N = 3SE +/- 0.00613, N = 3SE +/- 0.03254, N = 4SE +/- 0.01524, N = 3SE +/- 0.00856, N = 3SE +/- 0.02511, N = 3SE +/- 0.03176, N = 32.966922.671293.785313.051353.182063.099052.625244.660401. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: ResNet50 v1-12-int8 - Device: CPU - Executor: ParallelEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248880160240320400SE +/- 0.27, N = 3SE +/- 1.30, N = 3SE +/- 0.43, N = 3SE +/- 3.53, N = 4SE +/- 1.50, N = 3SE +/- 0.89, N = 3SE +/- 3.68, N = 3SE +/- 1.45, N = 3336.91374.23264.15327.75314.23322.61380.82214.551. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

SPECFEM3D

simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra. This test profile currently relies on CPU-based execution for SPECFEM3D and using a variety of their built-in examples/models for benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.1.1Model: Tomographic ModelEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881428425670SE +/- 0.32, N = 3SE +/- 0.34, N = 3SE +/- 0.26, N = 3SE +/- 0.25, N = 3SE +/- 0.53, N = 3SE +/- 0.48, N = 3SE +/- 0.20, N = 15SE +/- 0.06, N = 339.2828.0660.3852.2347.6949.5830.1352.891. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 2 - Resolution: 1080p - Samples Per Pixel: 16 - Renderer: Path Tracer - Acceleration: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248811K22K33K44K55KSE +/- 39.63, N = 3SE +/- 3.84, N = 3SE +/- 199.64, N = 3SE +/- 293.34, N = 3SE +/- 87.29, N = 3SE +/- 127.68, N = 3SE +/- 19.75, N = 3SE +/- 172.31, N = 32497218442503392553038690379001728542275

Zstd Compression

This test measures the time needed to compress/decompress a sample file (silesia.tar) using Zstd (Zstandard) compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 8, Long Mode - Compression SpeedEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24882004006008001000SE +/- 1.33, N = 3SE +/- 3.15, N = 3SE +/- 1.62, N = 3SE +/- 3.16, N = 3SE +/- 0.50, N = 3SE +/- 0.95, N = 3SE +/- 3.49, N = 3SE +/- 1.12, N = 3745.5961.2482.9888.7532.3527.9797.0504.11. (CC) gcc options: -O3 -pthread -lz -llzma

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Face Detection FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248830060090012001500SE +/- 0.16, N = 3SE +/- 0.56, N = 3SE +/- 0.60, N = 3SE +/- 0.35, N = 3SE +/- 0.62, N = 3SE +/- 0.52, N = 3SE +/- 1.73, N = 3SE +/- 8.08, N = 3647.10605.15761.78597.79579.07560.14590.801506.63MIN: 540.69 / MAX: 688.06MIN: 570.99 / MAX: 624.11MIN: 744.08 / MAX: 774.48MIN: 574.7 / MAX: 616.24MIN: 564.55 / MAX: 589.31MIN: 538 / MAX: 581.72MIN: 555.5 / MAX: 623.22MIN: 1180.01 / MAX: 1642.121. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Face Detection FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24883691215SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 39.2413.195.2510.006.907.1413.522.651. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

LuxCoreRender

LuxCoreRender is an open-source 3D physically based renderer formerly known as LuxRender. LuxCoreRender supports CPU-based rendering as well as GPU acceleration via OpenCL, NVIDIA CUDA, and NVIDIA OptiX interfaces. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: LuxCore Benchmark - Acceleration: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881.04632.09263.13894.18525.2315SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 33.124.551.683.122.332.414.651.69MIN: 1.19 / MAX: 3.62MIN: 1.97 / MAX: 5.16MIN: 0.5 / MAX: 2.06MIN: 1.17 / MAX: 3.62MIN: 0.78 / MAX: 2.76MIN: 0.82 / MAX: 2.85MIN: 2.03 / MAX: 5.27MIN: 0.51 / MAX: 2.04

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Face Detection FP16-INT8 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248890180270360450SE +/- 0.33, N = 3SE +/- 0.41, N = 3SE +/- 0.03, N = 3SE +/- 0.11, N = 3SE +/- 0.10, N = 3SE +/- 0.15, N = 3SE +/- 0.24, N = 3SE +/- 0.59, N = 3331.37312.52393.13313.09299.02285.97301.79384.09MIN: 179.64 / MAX: 354.6MIN: 167.86 / MAX: 330.96MIN: 378.57 / MAX: 400.92MIN: 291.27 / MAX: 322.77MIN: 288.59 / MAX: 303.77MIN: 150.08 / MAX: 303.22MIN: 278.87 / MAX: 327.8MIN: 285.45 / MAX: 543.491. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Face Detection FP16-INT8 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488612182430SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 318.0925.5410.1719.1213.3613.9726.4610.411. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 1 - Resolution: 1080p - Samples Per Pixel: 16 - Renderer: Path Tracer - Acceleration: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248811K22K33K44K55KSE +/- 16.70, N = 3SE +/- 12.67, N = 3SE +/- 80.90, N = 3SE +/- 280.33, N = 3SE +/- 55.41, N = 3SE +/- 43.80, N = 3SE +/- 33.74, N = 3SE +/- 112.38, N = 32471917460500552518638334374581711641514

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: GPT-2 - Device: CPU - Executor: ParallelEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488246810SE +/- 0.02651, N = 3SE +/- 0.04639, N = 3SE +/- 0.00330, N = 3SE +/- 0.02537, N = 3SE +/- 0.01600, N = 3SE +/- 0.02822, N = 3SE +/- 0.00809, N = 3SE +/- 0.03060, N = 38.192537.190058.404087.246398.242568.311078.151558.108641. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: GPT-2 - Device: CPU - Executor: ParallelEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.39, N = 3SE +/- 0.89, N = 3SE +/- 0.05, N = 3SE +/- 0.48, N = 3SE +/- 0.23, N = 3SE +/- 0.41, N = 3SE +/- 0.12, N = 3SE +/- 0.46, N = 3121.99139.02118.96137.92121.29120.28122.63123.281. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: SharpenEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881326395265SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 341572341303058361. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

LuxCoreRender

LuxCoreRender is an open-source 3D physically based renderer formerly known as LuxRender. LuxCoreRender supports CPU-based rendering as well as GPU acceleration via OpenCL, NVIDIA CUDA, and NVIDIA OptiX interfaces. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: DLSC - Acceleration: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881.1252.253.3754.55.625SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 33.484.882.063.472.672.755.002.13MIN: 3.39 / MAX: 3.64MIN: 4.77 / MAX: 5.16MIN: 2 / MAX: 2.12MIN: 3.38 / MAX: 3.64MIN: 2.59 / MAX: 2.76MIN: 2.66 / MAX: 2.86MIN: 4.85 / MAX: 5.35MIN: 2.03 / MAX: 2.65

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Machine Translation EN To DE FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.07, N = 3SE +/- 0.05, N = 3SE +/- 0.39, N = 3SE +/- 0.06, N = 3SE +/- 0.24, N = 3SE +/- 0.10, N = 3SE +/- 0.03, N = 3SE +/- 0.08, N = 358.5657.6863.2956.9853.4448.2658.43118.45MIN: 42.61 / MAX: 80.74MIN: 32.9 / MAX: 82.05MIN: 45.67 / MAX: 70.41MIN: 40.42 / MAX: 73.08MIN: 25.81 / MAX: 61.36MIN: 29.16 / MAX: 69.56MIN: 31.85 / MAX: 94.55MIN: 68.57 / MAX: 244.711. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Machine Translation EN To DE FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.13, N = 3SE +/- 0.12, N = 3SE +/- 0.39, N = 3SE +/- 0.10, N = 3SE +/- 0.34, N = 3SE +/- 0.17, N = 3SE +/- 0.06, N = 3SE +/- 0.02, N = 3102.35138.5063.15105.2074.7982.80136.7333.741. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: T5 Encoder - Device: CPU - Executor: ParallelEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488246810SE +/- 0.02105, N = 3SE +/- 0.01677, N = 3SE +/- 0.00546, N = 3SE +/- 0.01047, N = 3SE +/- 0.02627, N = 3SE +/- 0.06732, N = 3SE +/- 0.02831, N = 3SE +/- 0.03755, N = 36.414515.439506.789315.324537.226577.060976.378066.198761. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: T5 Encoder - Device: CPU - Executor: ParallelEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24884080120160200SE +/- 0.51, N = 3SE +/- 0.57, N = 3SE +/- 0.12, N = 3SE +/- 0.37, N = 3SE +/- 0.50, N = 3SE +/- 1.35, N = 3SE +/- 0.69, N = 3SE +/- 0.97, N = 3155.88183.82147.28187.78138.37141.63156.77161.311. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Person Detection FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.10, N = 3SE +/- 0.45, N = 3SE +/- 0.11, N = 3SE +/- 0.09, N = 3SE +/- 0.06, N = 3SE +/- 0.15, N = 3SE +/- 1.01, N = 3SE +/- 0.73, N = 381.5683.3090.5070.4476.3672.5289.98133.99MIN: 55.87 / MAX: 108.66MIN: 29.23 / MAX: 114.06MIN: 68.02 / MAX: 98.61MIN: 48.46 / MAX: 97.28MIN: 56.38 / MAX: 88.42MIN: 28.47 / MAX: 87.6MIN: 63.39 / MAX: 124.53MIN: 103.21 / MAX: 233.721. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Person Detection FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820406080100SE +/- 0.08, N = 3SE +/- 0.52, N = 3SE +/- 0.06, N = 3SE +/- 0.10, N = 3SE +/- 0.04, N = 3SE +/- 0.11, N = 3SE +/- 0.98, N = 3SE +/- 0.16, N = 373.5195.9544.1785.1152.3555.1188.8429.831. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Person Detection FP32 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.38, N = 3SE +/- 0.38, N = 3SE +/- 0.39, N = 3SE +/- 0.06, N = 3SE +/- 0.18, N = 3SE +/- 0.13, N = 3SE +/- 0.48, N = 3SE +/- 0.77, N = 380.7983.1890.0270.3176.5272.7090.49134.54MIN: 30.58 / MAX: 108.27MIN: 38.33 / MAX: 120MIN: 43.62 / MAX: 98.59MIN: 50.95 / MAX: 96.87MIN: 52.86 / MAX: 94.9MIN: 40.38 / MAX: 84.07MIN: 55.81 / MAX: 127.95MIN: 98.92 / MAX: 233.451. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Person Detection FP32 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820406080100SE +/- 0.35, N = 3SE +/- 0.44, N = 3SE +/- 0.20, N = 3SE +/- 0.07, N = 3SE +/- 0.13, N = 3SE +/- 0.09, N = 3SE +/- 0.47, N = 3SE +/- 0.17, N = 374.2296.1044.4085.2952.2454.9888.3229.711. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: EnhancedEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820406080100SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 367923864495094491. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Noise Suppression Poconet-Like FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248848121620SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 39.288.408.498.098.837.959.9015.21MIN: 5.38 / MAX: 30.18MIN: 5.53 / MAX: 31MIN: 5.12 / MAX: 20.17MIN: 5.45 / MAX: 16.52MIN: 5.3 / MAX: 14.21MIN: 5 / MAX: 23.85MIN: 6.53 / MAX: 37.69MIN: 10.08 / MAX: 64.691. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Noise Suppression Poconet-Like FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488400800120016002000SE +/- 0.77, N = 3SE +/- 2.01, N = 3SE +/- 1.27, N = 3SE +/- 1.93, N = 3SE +/- 1.25, N = 3SE +/- 0.86, N = 3SE +/- 1.23, N = 3SE +/- 1.17, N = 31274.271882.34697.691465.71890.99991.321587.74523.021. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Road Segmentation ADAS FP16-INT8 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488816243240SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.08, N = 3SE +/- 0.03, N = 3SE +/- 0.07, N = 3SE +/- 0.05, N = 317.0515.4521.3116.3416.8314.4816.7336.35MIN: 9.42 / MAX: 28.42MIN: 9.45 / MAX: 28.8MIN: 11.09 / MAX: 25.08MIN: 9.56 / MAX: 24.55MIN: 8.78 / MAX: 22.96MIN: 8.91 / MAX: 26.44MIN: 9.66 / MAX: 30.96MIN: 24.38 / MAX: 96.41. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Road Segmentation ADAS FP16-INT8 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488110220330440550SE +/- 0.40, N = 3SE +/- 0.32, N = 3SE +/- 0.12, N = 3SE +/- 0.69, N = 3SE +/- 1.13, N = 3SE +/- 0.50, N = 3SE +/- 1.94, N = 3SE +/- 0.28, N = 3351.23516.90187.28366.70237.30275.72476.63219.861. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Person Vehicle Bike Detection FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24883691215SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 35.645.106.605.325.044.785.129.94MIN: 3.48 / MAX: 18.22MIN: 3.41 / MAX: 14.16MIN: 4.29 / MAX: 16.53MIN: 3.77 / MAX: 11.07MIN: 3.47 / MAX: 17.02MIN: 3.18 / MAX: 16.31MIN: 3.61 / MAX: 15.43MIN: 6.2 / MAX: 41.651. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Person Vehicle Bike Detection FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248830060090012001500SE +/- 0.91, N = 3SE +/- 2.21, N = 3SE +/- 2.84, N = 3SE +/- 5.46, N = 3SE +/- 5.03, N = 3SE +/- 0.80, N = 3SE +/- 2.52, N = 3SE +/- 0.87, N = 31059.991560.99603.921124.06790.63833.731556.07401.421. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: Noise-GaussianEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.33, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.58, N = 3SE +/- 1.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 38911560877376119701. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Road Segmentation ADAS FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488918273645SE +/- 0.05, N = 3SE +/- 0.07, N = 3SE +/- 0.06, N = 3SE +/- 0.07, N = 3SE +/- 0.12, N = 3SE +/- 0.09, N = 3SE +/- 0.04, N = 3SE +/- 0.09, N = 319.6818.3227.2317.7623.2717.7925.1539.48MIN: 10.8 / MAX: 33.75MIN: 10.07 / MAX: 36.96MIN: 12.7 / MAX: 33.19MIN: 11.78 / MAX: 28.07MIN: 8.9 / MAX: 35.88MIN: 9.38 / MAX: 30.36MIN: 11.63 / MAX: 50.5MIN: 22.02 / MAX: 90.71. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Road Segmentation ADAS FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248890180270360450SE +/- 0.76, N = 3SE +/- 1.74, N = 3SE +/- 0.31, N = 3SE +/- 1.33, N = 3SE +/- 0.84, N = 3SE +/- 1.10, N = 3SE +/- 0.46, N = 3SE +/- 0.22, N = 3304.22435.77146.62337.45171.63224.43317.25101.211. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Person Re-Identification Retail FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248848121620SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.11, N = 34.644.285.154.364.073.874.1118.22MIN: 2.6 / MAX: 15.94MIN: 2.63 / MAX: 16.2MIN: 3.03 / MAX: 14.81MIN: 2.55 / MAX: 15.23MIN: 2.41 / MAX: 9.33MIN: 2.38 / MAX: 13.89MIN: 2.45 / MAX: 41.27MIN: 10.36 / MAX: 177.581. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Person Re-Identification Retail FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488400800120016002000SE +/- 0.33, N = 3SE +/- 0.50, N = 3SE +/- 0.80, N = 3SE +/- 1.33, N = 3SE +/- 1.30, N = 3SE +/- 0.99, N = 3SE +/- 0.22, N = 3SE +/- 2.58, N = 31286.311863.36773.961372.19978.601029.881938.32437.961. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Handwritten English Recognition FP16-INT8 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881020304050SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.09, N = 3SE +/- 0.07, N = 3SE +/- 0.04, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.12, N = 322.1120.5719.8620.3120.2219.1320.0945.70MIN: 14.68 / MAX: 40.04MIN: 14.62 / MAX: 37.71MIN: 13.85 / MAX: 30.49MIN: 14.76 / MAX: 29.79MIN: 14.54 / MAX: 30.69MIN: 12.57 / MAX: 34MIN: 13.66 / MAX: 38.02MIN: 32.92 / MAX: 117.921. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Handwritten English Recognition FP16-INT8 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24882004006008001000SE +/- 0.61, N = 3SE +/- 0.33, N = 3SE +/- 1.40, N = 3SE +/- 1.97, N = 3SE +/- 0.81, N = 3SE +/- 0.59, N = 3SE +/- 1.22, N = 3SE +/- 0.45, N = 3541.92776.84301.93589.77395.19417.64795.05174.921. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Vehicle Detection FP16-INT8 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248848121620SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.05, N = 35.335.016.195.024.754.534.8215.95MIN: 2.88 / MAX: 17.76MIN: 2.78 / MAX: 17.17MIN: 3.5 / MAX: 13.93MIN: 2.86 / MAX: 11.09MIN: 2.76 / MAX: 10.38MIN: 2.65 / MAX: 17.47MIN: 2.78 / MAX: 17.83MIN: 9.85 / MAX: 70.281. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Vehicle Detection FP16-INT8 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488400800120016002000SE +/- 0.37, N = 3SE +/- 0.49, N = 3SE +/- 1.34, N = 3SE +/- 1.23, N = 3SE +/- 2.16, N = 3SE +/- 0.43, N = 3SE +/- 1.19, N = 3SE +/- 1.52, N = 31120.771588.30644.501191.72838.46879.501653.69500.861. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Handwritten English Recognition FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881224364860SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.30, N = 3SE +/- 0.08, N = 3SE +/- 0.20, N = 3SE +/- 0.05, N = 3SE +/- 0.09, N = 3SE +/- 0.12, N = 323.4721.8522.0121.4021.7020.3721.8953.07MIN: 17.04 / MAX: 41.98MIN: 15.94 / MAX: 38.89MIN: 13.52 / MAX: 35.51MIN: 14.65 / MAX: 31.23MIN: 16.33 / MAX: 33.88MIN: 14.74 / MAX: 35.54MIN: 14.64 / MAX: 40.37MIN: 35.4 / MAX: 130.321. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Handwritten English Recognition FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488160320480640800SE +/- 0.66, N = 3SE +/- 0.36, N = 3SE +/- 3.75, N = 3SE +/- 1.93, N = 3SE +/- 3.35, N = 3SE +/- 1.00, N = 3SE +/- 3.18, N = 3SE +/- 0.33, N = 3510.62731.48272.46559.88368.26392.23729.84150.641. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Face Detection Retail FP16-INT8 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881.14082.28163.42244.56325.704SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 33.643.513.273.533.353.163.335.07MIN: 1.95 / MAX: 15.55MIN: 1.92 / MAX: 13.78MIN: 1.84 / MAX: 8.95MIN: 1.95 / MAX: 9.62MIN: 1.81 / MAX: 10.37MIN: 1.77 / MAX: 13.91MIN: 1.88 / MAX: 15.67MIN: 2.36 / MAX: 27.991. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Face Detection Retail FP16-INT8 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248810002000300040005000SE +/- 1.50, N = 3SE +/- 3.97, N = 3SE +/- 0.53, N = 3SE +/- 2.70, N = 3SE +/- 2.13, N = 3SE +/- 1.63, N = 3SE +/- 1.99, N = 3SE +/- 3.53, N = 33221.204486.581804.903338.382329.232495.554734.271559.061. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

SPECFEM3D

simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra. This test profile currently relies on CPU-based execution for SPECFEM3D and using a variety of their built-in examples/models for benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.1.1Model: Homogeneous HalfspaceEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820406080100SE +/- 0.57, N = 3SE +/- 0.48, N = 3SE +/- 1.09, N = 3SE +/- 0.43, N = 3SE +/- 0.54, N = 3SE +/- 0.35, N = 3SE +/- 0.28, N = 3SE +/- 0.41, N = 350.7036.3177.2367.5659.6363.8739.5667.111. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24880.090.180.270.360.45SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.320.300.320.310.290.280.290.40MIN: 0.16 / MAX: 9.37MIN: 0.16 / MAX: 8.01MIN: 0.16 / MAX: 7.47MIN: 0.16 / MAX: 7.6MIN: 0.16 / MAX: 5.09MIN: 0.15 / MAX: 7.71MIN: 0.16 / MAX: 8.63MIN: 0.18 / MAX: 8.351. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248811K22K33K44K55KSE +/- 68.04, N = 3SE +/- 32.19, N = 3SE +/- 103.84, N = 3SE +/- 42.38, N = 3SE +/- 5.77, N = 3SE +/- 19.90, N = 3SE +/- 47.80, N = 3SE +/- 38.88, N = 333784.7446506.7517920.7535169.3125613.2926230.8549485.3518507.311. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Vehicle Detection FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248848121620SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.06, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.01, N = 38.787.8910.947.759.138.4411.3616.04MIN: 4.89 / MAX: 24.3MIN: 4.2 / MAX: 25.04MIN: 4.95 / MAX: 20.28MIN: 4.8 / MAX: 13.5MIN: 4.28 / MAX: 20.17MIN: 4.55 / MAX: 19.23MIN: 4.12 / MAX: 36.81MIN: 9.49 / MAX: 56.981. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Vehicle Detection FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24882004006008001000SE +/- 0.50, N = 3SE +/- 2.08, N = 3SE +/- 2.00, N = 3SE +/- 1.18, N = 3SE +/- 0.47, N = 3SE +/- 1.08, N = 3SE +/- 2.53, N = 3SE +/- 0.14, N = 3681.131011.61364.22772.21436.41472.57701.84249.021. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Weld Porosity Detection FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488714212835SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.06, N = 312.8212.0411.2711.9111.5111.0511.7230.05MIN: 6.57 / MAX: 26.61MIN: 6.3 / MAX: 25.96MIN: 5.84 / MAX: 21.18MIN: 6.32 / MAX: 23.15MIN: 5.95 / MAX: 21.47MIN: 5.83 / MAX: 23.56MIN: 6 / MAX: 26.83MIN: 17.38 / MAX: 87.991. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Weld Porosity Detection FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248830060090012001500SE +/- 0.30, N = 3SE +/- 0.25, N = 3SE +/- 0.14, N = 3SE +/- 0.16, N = 3SE +/- 0.27, N = 3SE +/- 0.13, N = 3SE +/- 1.37, N = 3SE +/- 0.54, N = 3934.241326.24531.481005.38693.78723.131362.00265.831. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: SwirlEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248850100150200250SE +/- 0.33, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 31812431061631371392501301. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Age Gender Recognition Retail 0013 FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24880.2250.450.6750.91.125SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.460.430.410.440.410.400.411.00MIN: 0.22 / MAX: 9.38MIN: 0.22 / MAX: 7.53MIN: 0.22 / MAX: 5.13MIN: 0.22 / MAX: 7.12MIN: 0.21 / MAX: 8.32MIN: 0.2 / MAX: 7.69MIN: 0.21 / MAX: 10.02MIN: 0.4 / MAX: 10.731. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Age Gender Recognition Retail 0013 FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24887K14K21K28K35KSE +/- 16.45, N = 3SE +/- 29.98, N = 3SE +/- 1.45, N = 3SE +/- 16.68, N = 3SE +/- 12.32, N = 3SE +/- 8.32, N = 3SE +/- 16.51, N = 3SE +/- 31.71, N = 323774.8833111.8214005.4625279.9718120.3818615.1234815.397499.471. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Weld Porosity Detection FP16-INT8 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488246810SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 36.496.215.756.175.875.635.967.78MIN: 3.31 / MAX: 18.52MIN: 3.24 / MAX: 18.16MIN: 3.02 / MAX: 14.8MIN: 3.21 / MAX: 15.44MIN: 3.03 / MAX: 15.22MIN: 2.98 / MAX: 16.09MIN: 3.08 / MAX: 18.52MIN: 3.64 / MAX: 44.071. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Weld Porosity Detection FP16-INT8 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24886001200180024003000SE +/- 0.27, N = 3SE +/- 0.50, N = 3SE +/- 0.80, N = 3SE +/- 1.19, N = 3SE +/- 0.44, N = 3SE +/- 0.60, N = 3SE +/- 4.11, N = 3SE +/- 2.31, N = 31835.782563.791040.151936.551356.641417.332672.511022.101. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Face Detection Retail FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24883691215SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.02, N = 32.702.512.982.622.372.212.379.76MIN: 1.33 / MAX: 14.55MIN: 1.32 / MAX: 13.29MIN: 1.64 / MAX: 9.56MIN: 1.36 / MAX: 13.53MIN: 1.25 / MAX: 6.4MIN: 1.23 / MAX: 13.12MIN: 1.27 / MAX: 12.75MIN: 4.34 / MAX: 46.891. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Face Detection Retail FP16 - Device: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24887001400210028003500SE +/- 0.82, N = 3SE +/- 1.96, N = 3SE +/- 3.43, N = 3SE +/- 2.38, N = 3SE +/- 1.26, N = 3SE +/- 0.66, N = 3SE +/- 6.25, N = 3SE +/- 1.96, N = 32184.943099.901330.812254.381659.061781.623269.12816.171. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

John The Ripper

This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: MD5EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488900K1800K2700K3600K4500KSE +/- 8082.90, N = 3SE +/- 17559.42, N = 3SE +/- 17071.74, N = 3SE +/- 333.33, N = 3SE +/- 666.67, N = 3SE +/- 7423.69, N = 3SE +/- 11590.23, N = 3SE +/- 2504.83, N = 328070003825000156833330303332219333223433341180009197911. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2

Speedb

Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Read Random Write RandomEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488600K1200K1800K2400K3000KSE +/- 6109.05, N = 3SE +/- 1095.04, N = 3SE +/- 2410.81, N = 3SE +/- 284.11, N = 3SE +/- 1251.53, N = 3SE +/- 1406.43, N = 3SE +/- 2327.92, N = 3SE +/- 4914.43, N = 3216669927063851806410184197022225562138776277019521209151. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: HWB Color SpaceEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248870140210280350SE +/- 1.53, N = 3SE +/- 1.73, N = 3SE +/- 0.58, N = 3SE +/- 0.88, N = 3SE +/- 1.20, N = 3SE +/- 0.33, N = 3SE +/- 0.58, N = 3SE +/- 0.67, N = 32582682272362492552783151. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: ResizingEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248870140210280350SE +/- 0.88, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 1.00, N = 3SE +/- 1.86, N = 32843252022582502553382441. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

Speedb

Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Random ReadEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248830M60M90M120M150MSE +/- 36870.14, N = 3SE +/- 806005.16, N = 3SE +/- 43642.52, N = 3SE +/- 202059.67, N = 3SE +/- 63522.73, N = 3SE +/- 34435.41, N = 3SE +/- 286737.26, N = 3SE +/- 170759.02, N = 39475027913284399253146372920064216944301171648683134971626578244261. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

RocksDB

This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Read Random Write RandomEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488600K1200K1800K2400K3000KSE +/- 16096.70, N = 3SE +/- 8693.67, N = 3SE +/- 24161.52, N = 3SE +/- 1307.98, N = 3SE +/- 1649.04, N = 3SE +/- 5891.62, N = 3SE +/- 24549.22, N = 3SE +/- 17144.87, N = 3227855629295611781795181777522474812131122291575819765331. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Read While WritingEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488900K1800K2700K3600K4500KSE +/- 10561.84, N = 3SE +/- 23577.88, N = 3SE +/- 20912.63, N = 3SE +/- 24911.17, N = 3SE +/- 6544.07, N = 3SE +/- 22460.23, N = 3SE +/- 7045.87, N = 3SE +/- 3646.67, N = 3313093443345931865521300361624898542519594429463119427601. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: super-resolution-10 - Device: CPU - Executor: ParallelEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248848121620SE +/- 0.01982, N = 3SE +/- 0.03535, N = 3SE +/- 0.19593, N = 3SE +/- 0.04762, N = 3SE +/- 0.02640, N = 3SE +/- 0.02135, N = 3SE +/- 0.06152, N = 3SE +/- 0.09981, N = 39.745918.3085415.5584010.5013011.2108010.918207.5008013.179201. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: super-resolution-10 - Device: CPU - Executor: ParallelEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.21, N = 3SE +/- 0.51, N = 3SE +/- 0.80, N = 3SE +/- 0.43, N = 3SE +/- 0.21, N = 3SE +/- 0.18, N = 3SE +/- 1.10, N = 3SE +/- 0.57, N = 3102.60120.3564.2995.2289.2091.59133.3275.881. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

John The Ripper

This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: HMAC-SHA512EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248840M80M120M160M200MSE +/- 1126651.23, N = 3SE +/- 243467.31, N = 3SE +/- 24875.24, N = 3SE +/- 58043.09, N = 3SE +/- 110045.95, N = 3SE +/- 411753.30, N = 3SE +/- 580393.26, N = 3SE +/- 518797.01, N = 313634400017090800089217667144557000117072667116770333181331000407210001. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2

RocksDB

This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Random ReadEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248830M60M90M120M150MSE +/- 74109.11, N = 3SE +/- 276158.53, N = 3SE +/- 45665.92, N = 3SE +/- 342211.57, N = 3SE +/- 68428.53, N = 3SE +/- 144351.01, N = 3SE +/- 780691.20, N = 3SE +/- 236765.68, N = 39172694213000673751334146891950286795936069408254130693063475122601. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

libavif avifenc

This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 1.0Encoder Speed: 2EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881428425670SE +/- 0.18, N = 3SE +/- 0.33, N = 7SE +/- 0.23, N = 3SE +/- 0.56, N = 4SE +/- 0.03, N = 3SE +/- 0.13, N = 3SE +/- 0.03, N = 3SE +/- 0.18, N = 344.3437.1164.1445.7352.5651.9436.3361.841. (CXX) g++ options: -O3 -fPIC -lm

Xcompact3d Incompact3d

Xcompact3d Incompact3d is a Fortran-MPI based, finite difference high-performance code for solving the incompressible Navier-Stokes equation and as many as you need scalar transport equations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterXcompact3d Incompact3d 2021-03-11Input: input.i3d 129 Cells Per DirectionEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488510152025SE +/- 0.17, N = 15SE +/- 0.11, N = 15SE +/- 0.02, N = 3SE +/- 0.15, N = 15SE +/- 0.03, N = 3SE +/- 0.18, N = 9SE +/- 0.12, N = 15SE +/- 0.22, N = 317.2913.8221.8014.7620.3921.8716.5919.751. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

Intel Open Image Denoise

Open Image Denoise is a denoising library for ray-tracing and part of the Intel oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 2.2Run: RT.hdr_alb_nrm.3840x2160 - Device: CPU-OnlyEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24880.19350.3870.58050.7740.9675SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.640.840.360.640.470.480.860.42

OpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 2.2Run: RT.ldr_alb_nrm.3840x2160 - Device: CPU-OnlyEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24880.19350.3870.58050.7740.9675SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.640.840.360.640.470.480.860.42

Z3 Theorem Prover

The Z3 Theorem Prover / SMT solver is developed by Microsoft Research under the MIT license. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterZ3 Theorem Prover 4.12.1SMT File: 2.smt2EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881428425670SE +/- 0.39, N = 3SE +/- 0.02, N = 3SE +/- 0.33, N = 3SE +/- 0.51, N = 3SE +/- 0.40, N = 3SE +/- 0.27, N = 3SE +/- 0.53, N = 3SE +/- 0.51, N = 350.4647.5749.3247.2347.7348.6848.8863.081. (CXX) g++ options: -lpthread -std=c++17 -fvisibility=hidden -mfpmath=sse -msse -msse2 -O3 -fPIC

Timed PHP Compilation

This test times how long it takes to build PHP. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed PHP Compilation 8.3.4Time To CompileEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881428425670SE +/- 0.24, N = 3SE +/- 0.22, N = 3SE +/- 0.03, N = 3SE +/- 0.21, N = 3SE +/- 0.07, N = 3SE +/- 0.10, N = 3SE +/- 0.18, N = 3SE +/- 0.20, N = 345.5940.1063.3043.9954.0453.7640.1154.45

PyTorch

This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 1 - Model: ResNet-50EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881530456075SE +/- 0.11, N = 3SE +/- 0.68, N = 6SE +/- 0.26, N = 3SE +/- 0.30, N = 3SE +/- 0.65, N = 3SE +/- 1.84, N = 15SE +/- 0.90, N = 15SE +/- 1.37, N = 1256.9668.8058.4961.3468.6767.4865.6961.49MIN: 52.49 / MAX: 58MIN: 59.18 / MAX: 72.68MIN: 55.22 / MAX: 59.43MIN: 50.53 / MAX: 62.03MIN: 64.12 / MAX: 71.37MIN: 46.76 / MAX: 74.15MIN: 53.12 / MAX: 69.76MIN: 40.06 / MAX: 72.22

m-queens

A solver for the N-queens problem with multi-threading support via the OpenMP library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterm-queens 1.2Time To SolveEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881632486480SE +/- 0.05, N = 3SE +/- 0.03, N = 3SE +/- 0.08, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.62, N = 339.6528.5371.9840.4254.3953.5027.9172.851. (CXX) g++ options: -fopenmp -O2 -march=native

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs (and GPUs via SYCL) and supporting instruction sets such as SSE, AVX, AVX2, and AVX-512. Embree also supports making use of the Intel SPMD Program Compiler (ISPC). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 4.3Binary: Pathtracer ISPC - Model: Asian Dragon ObjEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488714212835SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.10, N = 321.2429.1712.2021.7815.6415.6927.3914.55MIN: 21.05 / MAX: 21.73MIN: 28.89 / MAX: 30.01MIN: 12.12 / MAX: 12.41MIN: 21.53 / MAX: 22.24MIN: 15.52 / MAX: 15.91MIN: 15.56 / MAX: 16.03MIN: 27.11 / MAX: 28.25MIN: 14.14 / MAX: 16.68

Coremark

This is a test of EEMBC CoreMark processor benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations/Sec, More Is BetterCoremark 1.0CoreMark Size 666 - Iterations Per SecondEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488200K400K600K800K1000KSE +/- 7794.37, N = 3SE +/- 1179.52, N = 3SE +/- 954.28, N = 3SE +/- 7061.60, N = 3SE +/- 4350.78, N = 3SE +/- 1076.82, N = 3SE +/- 9073.63, N = 3SE +/- 3256.58, N = 15648144.00923331.48372162.82625696.15491021.69499103.50951674.51430017.461. (CC) gcc options: -O2 -lrt" -lrt

Renaissance

Renaissance is a suite of benchmarks designed to test the Java JVM from Apache Spark to a Twitter-like service to Scala and other features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.14Test: Savina Reactors.IOEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24888001600240032004000SE +/- 30.89, N = 3SE +/- 41.68, N = 3SE +/- 12.81, N = 3SE +/- 38.91, N = 4SE +/- 13.60, N = 3SE +/- 28.17, N = 3SE +/- 15.03, N = 3SE +/- 16.31, N = 33338.23350.93288.43515.93060.53042.03265.23695.4MIN: 3297.53 / MAX: 4743.03MIN: 3285.69 / MAX: 4786.32MIN: 3264.51 / MAX: 4523.94MIN: 3407.81 / MAX: 4967.72MIN: 3036.07 / MAX: 4235.11MIN: 3011.87 / MAX: 4321.7MIN: 3236.75 / MAX: 4741.82MIN: 3664.8 / MAX: 5363.87

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: MediumEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488510152025SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.05, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 314.6620.659.5914.4012.1012.4922.029.281. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

SVT-HEVC

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 1 - Input: Bosphorus 1080pEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488612182430SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 315.7122.119.1215.8911.9012.2823.109.161. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt

VVenC

VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVVenC 1.11Video Input: Bosphorus 4K - Video Preset: FasterEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248848121620SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.05, N = 314.5416.8210.4914.6012.8013.1417.4411.771. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects

x265

This is a simple test of the x265 encoder run on the CPU with 1080p and 4K options for H.265 video encode performance with x265. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betterx265 3.4Video Input: Bosphorus 4KEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488816243240SE +/- 0.16, N = 3SE +/- 0.20, N = 3SE +/- 0.19, N = 3SE +/- 0.11, N = 3SE +/- 0.22, N = 3SE +/- 0.15, N = 15SE +/- 0.08, N = 3SE +/- 0.17, N = 327.4334.0217.6027.2016.7717.6134.6519.481. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

FFmpeg

This is a benchmark of the FFmpeg multimedia framework. The FFmpeg test profile is making use of a modified version of vbench from Columbia University's Architecture and Design Lab (ARCADE) [http://arcade.cs.columbia.edu/vbench/] that is a benchmark for video-as-a-service workloads. The test profile offers the options of a range of vbench scenarios based on freely distributable video content and offers the options of using the x264 or x265 video encoders for transcoding. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 6.1Encoder: libx265 - Scenario: LiveEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24884080120160200SE +/- 0.29, N = 3SE +/- 0.59, N = 3SE +/- 0.31, N = 3SE +/- 1.34, N = 3SE +/- 0.58, N = 3SE +/- 0.17, N = 3SE +/- 0.71, N = 3SE +/- 1.39, N = 3177.93176.69162.38171.60176.63179.99186.42178.151. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

JPEG-XL libjxl

The JPEG XL Image Coding System is designed to provide next-generation JPEG image capabilities with JPEG XL offering better image quality and compression over legacy JPEG. This test profile is currently focused on the multi-threaded JPEG XL image encode performance using the reference libjxl library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterJPEG-XL libjxl 0.10.1Input: JPEG - Quality: 90EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881122334455SE +/- 0.46, N = 3SE +/- 0.47, N = 6SE +/- 0.13, N = 3SE +/- 0.38, N = 3SE +/- 0.01, N = 3SE +/- 0.30, N = 3SE +/- 0.37, N = 10SE +/- 0.09, N = 341.7747.0530.1940.0135.6037.1848.5932.511. (CXX) g++ options: -fno-rtti -O3 -fPIE -pie -lm

OpenBenchmarking.orgMP/s, More Is BetterJPEG-XL libjxl 0.10.1Input: PNG - Quality: 90EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881122334455SE +/- 0.44, N = 3SE +/- 0.24, N = 3SE +/- 0.08, N = 3SE +/- 0.30, N = 10SE +/- 0.16, N = 3SE +/- 0.36, N = 3SE +/- 0.28, N = 3SE +/- 0.19, N = 340.1845.3829.6538.7235.0836.0248.5331.811. (CXX) g++ options: -fno-rtti -O3 -fPIE -pie -lm

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs (and GPUs via SYCL) and supporting instruction sets such as SSE, AVX, AVX2, and AVX-512. Embree also supports making use of the Intel SPMD Program Compiler (ISPC). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 4.3Binary: Pathtracer ISPC - Model: CrownEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488714212835SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 321.9932.1812.3322.5716.0916.4631.3713.39MIN: 21.74 / MAX: 22.75MIN: 31.73 / MAX: 33MIN: 12.21 / MAX: 12.61MIN: 22.22 / MAX: 23.17MIN: 15.93 / MAX: 16.53MIN: 16.24 / MAX: 16.97MIN: 30.96 / MAX: 32.3MIN: 13.27 / MAX: 16.49

QuantLib

QuantLib is an open-source library/framework around quantitative finance for modeling, trading and risk management scenarios. QuantLib is written in C++ with Boost and its built-in benchmark used reports the QuantLib Benchmark Index benchmark score. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMFLOPS, More Is BetterQuantLib 1.32Configuration: Multi-ThreadedEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820K40K60K80K100KSE +/- 164.89, N = 3SE +/- 255.88, N = 3SE +/- 41.80, N = 3SE +/- 315.31, N = 3SE +/- 12.44, N = 3SE +/- 49.82, N = 3SE +/- 169.22, N = 3SE +/- 294.52, N = 356873.280769.432626.857419.442668.643490.783365.137214.41. (CXX) g++ options: -O3 -march=native -fPIE -pie

Timed FFmpeg Compilation

This test times how long it takes to build the FFmpeg multimedia library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed FFmpeg Compilation 6.1Time To CompileEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881224364860SE +/- 0.17, N = 3SE +/- 0.06, N = 3SE +/- 0.13, N = 3SE +/- 0.03, N = 3SE +/- 0.06, N = 3SE +/- 0.02, N = 3SE +/- 0.12, N = 3SE +/- 0.33, N = 333.0325.2553.8830.5542.9442.2025.3050.05

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 32 - Model: GoogLeNetEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.06, N = 3SE +/- 0.05, N = 3SE +/- 0.01, N = 3SE +/- 0.13, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.14, N = 3SE +/- 0.10, N = 3111.99139.1174.93115.4892.8593.71135.3078.79

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: SP.BEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24885K10K15K20K25KSE +/- 246.27, N = 15SE +/- 245.10, N = 4SE +/- 2.11, N = 3SE +/- 51.70, N = 4SE +/- 9.02, N = 3SE +/- 12.84, N = 3SE +/- 253.63, N = 3SE +/- 44.80, N = 317215.9224785.6511155.6425252.3511203.8311152.3118660.9012750.971. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs (and GPUs via SYCL) and supporting instruction sets such as SSE, AVX, AVX2, and AVX-512. Embree also supports making use of the Intel SPMD Program Compiler (ISPC). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 4.3Binary: Pathtracer ISPC - Model: Asian DragonEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488816243240SE +/- 0.02, N = 3SE +/- 0.06, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.12, N = 325.0134.2414.3625.5318.3918.4532.2216.58MIN: 24.79 / MAX: 25.66MIN: 33.88 / MAX: 35.24MIN: 14.27 / MAX: 14.68MIN: 25.23 / MAX: 26.17MIN: 18.28 / MAX: 18.74MIN: 18.3 / MAX: 18.87MIN: 31.88 / MAX: 33.23MIN: 16.29 / MAX: 19.34

Timed Wasmer Compilation

This test times how long it takes to compile Wasmer. Wasmer is written in the Rust programming language and is a WebAssembly runtime implementation that supports WASI and EmScripten. This test profile builds Wasmer with the Cranelift and Singlepast compiler features enabled. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Wasmer Compilation 2.3Time To CompileEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881020304050SE +/- 0.36, N = 3SE +/- 0.25, N = 3SE +/- 0.12, N = 3SE +/- 0.14, N = 3SE +/- 0.14, N = 3SE +/- 0.23, N = 3SE +/- 0.13, N = 3SE +/- 0.23, N = 332.5629.5042.1130.4736.9036.9029.4438.131. (CC) gcc options: -m64 -ldl -lgcc_s -lutil -lrt -lpthread -lm -lc -pie -nodefaultlibs

SVT-VP9

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-VP9 CPU-based multi-threaded video encoder for the VP9 video format with a sample YUV input video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: Visual Quality Optimized - Input: Bosphorus 4KEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820406080100SE +/- 0.79, N = 15SE +/- 1.39, N = 15SE +/- 0.51, N = 5SE +/- 0.75, N = 15SE +/- 0.64, N = 5SE +/- 0.56, N = 5SE +/- 1.42, N = 15SE +/- 0.47, N = 1580.93100.0550.5180.8760.9761.90100.0059.561. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

srsRAN Project

srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.10.1-20240325Test: PUSCH Processor Benchmark, Throughput ThreadEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24884080120160200SE +/- 0.00, N = 4SE +/- 0.00, N = 4SE +/- 3.58, N = 15SE +/- 0.03, N = 4SE +/- 0.00, N = 4SE +/- 0.00, N = 4SE +/- 0.00, N = 4SE +/- 7.12, N = 15183.6183.6179.8183.4183.4183.6183.5153.6-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 106-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 106-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 105.9 / MAX: 183.4-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 105.8-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 105.9-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 106-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 105.9MIN: 105.1 / MAX: 182.11. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -fno-trapping-math -fno-math-errno -ldl

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 64 - Model: AlexNetEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248870140210280350SE +/- 0.15, N = 3SE +/- 0.33, N = 3SE +/- 0.07, N = 3SE +/- 0.15, N = 3SE +/- 0.08, N = 3SE +/- 0.21, N = 3SE +/- 0.34, N = 3SE +/- 0.21, N = 3256.65315.45160.28248.73204.83206.40322.53173.78

miniBUDE

MiniBUDE is a mini application for the the core computation of the Bristol University Docking Engine (BUDE). This test profile currently makes use of the OpenMP implementation of miniBUDE for CPU benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBillion Interactions/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM1EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881122334455SE +/- 0.10, N = 3SE +/- 0.17, N = 4SE +/- 0.00, N = 3SE +/- 0.26, N = 3SE +/- 0.01, N = 3SE +/- 0.10, N = 3SE +/- 0.03, N = 4SE +/- 0.06, N = 334.6048.8219.0930.6825.8526.2949.9512.451. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

OpenBenchmarking.orgGFInst/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM1EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248830060090012001500SE +/- 2.41, N = 3SE +/- 4.13, N = 4SE +/- 0.09, N = 3SE +/- 6.54, N = 3SE +/- 0.15, N = 3SE +/- 2.56, N = 3SE +/- 0.84, N = 4SE +/- 1.56, N = 3864.961220.53477.36766.92646.26657.351248.81311.311. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

VVenC

VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVVenC 1.11Video Input: Bosphorus 1080p - Video Preset: FastEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488612182430SE +/- 0.08, N = 3SE +/- 0.10, N = 3SE +/- 0.02, N = 3SE +/- 0.10, N = 3SE +/- 0.06, N = 3SE +/- 0.08, N = 3SE +/- 0.07, N = 3SE +/- 0.06, N = 320.2121.7215.5419.9418.3018.7723.1915.461. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects

Aircrack-ng

Aircrack-ng is a tool for assessing WiFi/WLAN network security. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgk/s, More Is BetterAircrack-ng 1.7EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820K40K60K80K100KSE +/- 7.31, N = 3SE +/- 75.15, N = 3SE +/- 28.81, N = 3SE +/- 41.76, N = 3SE +/- 20.41, N = 3SE +/- 45.75, N = 3SE +/- 153.97, N = 3SE +/- 17.95, N = 362286.2088626.4535520.7367324.6846525.7848527.8292259.9535441.711. (CXX) g++ options: -std=gnu++17 -O3 -fvisibility=hidden -fcommon -rdynamic -lnl-3 -lnl-genl-3 -lpcre -lsqlite3 -lpthread -lz -lssl -lcrypto -lhwloc -ldl -lm -pthread

John The Ripper

This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: WPA PSKEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248830K60K90K120K150KSE +/- 646.01, N = 3SE +/- 585.52, N = 3SE +/- 52.29, N = 3SE +/- 115.44, N = 3SE +/- 166.40, N = 3SE +/- 221.78, N = 3SE +/- 425.10, N = 3SE +/- 135.70, N = 3112461155044635661201398419785532160664295601. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: bcryptEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24889K18K27K36K45KSE +/- 157.44, N = 3SE +/- 89.13, N = 3SE +/- 12.53, N = 3SE +/- 12.90, N = 3SE +/- 7.88, N = 3SE +/- 43.32, N = 3SE +/- 104.20, N = 3SE +/- 127.68, N = 331119419611765432882234072323543695187871. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: BlowfishEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24889K18K27K36K45KSE +/- 156.57, N = 3SE +/- 175.99, N = 3SE +/- 17.13, N = 3SE +/- 7.77, N = 3SE +/- 3.00, N = 3SE +/- 44.47, N = 3SE +/- 39.87, N = 3SE +/- 68.54, N = 331440418891762132864234032324343795188411. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 16 - Buffer Length: 256 - Filter Length: 512EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248860M120M180M240M300MSE +/- 1274497.72, N = 3SE +/- 2913213.19, N = 3SE +/- 138604.15, N = 3SE +/- 168358.08, N = 3SE +/- 170098.01, N = 3SE +/- 75055.53, N = 3SE +/- 520000.00, N = 3SE +/- 1245404.89, N = 32335966672825633331564333332301533332062700002116700002923900001592700001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Cpuminer-Opt

Cpuminer-Opt is a fork of cpuminer-multi that carries a wide range of CPU performance optimizations for measuring the potential cryptocurrency mining performance of the CPU/processor with a wide variety of cryptocurrencies. The benchmark reports the hash speed for the CPU mining performance for the selected cryptocurrency. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: LBC, LBRY CreditsEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24883K6K9K12K15KSE +/- 3.33, N = 3SE +/- 106.67, N = 3SE +/- 2.97, N = 3SE +/- 3.33, N = 3SE +/- 27.28, N = 3SE +/- 7.26, N = 3SE +/- 5.77, N = 3SE +/- 15.18, N = 311623.0015777.006265.1011867.008323.698389.1516120.006037.261. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

Stress-NG

Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Wide Vector MathEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488200K400K600K800K1000KSE +/- 1009.32, N = 3SE +/- 875.08, N = 3SE +/- 31.25, N = 3SE +/- 164.21, N = 3SE +/- 84.63, N = 3SE +/- 205.90, N = 3SE +/- 1811.04, N = 3SE +/- 373.64, N = 3539442.24749237.41307368.19568714.45402835.32413926.82778238.33576480.151. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -ljpeg -lmpfr -lpthread -lrt -lsctp -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Vector Floating PointEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248813K26K39K52K65KSE +/- 284.33, N = 3SE +/- 280.00, N = 3SE +/- 108.35, N = 3SE +/- 59.97, N = 3SE +/- 129.67, N = 3SE +/- 415.80, N = 3SE +/- 261.75, N = 3SE +/- 378.33, N = 341241.4556703.0620773.7041113.8528866.8830936.5059037.1428846.061. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -ljpeg -lmpfr -lpthread -lrt -lsctp -lz

Cpuminer-Opt

Cpuminer-Opt is a fork of cpuminer-multi that carries a wide range of CPU performance optimizations for measuring the potential cryptocurrency mining performance of the CPU/processor with a wide variety of cryptocurrencies. The benchmark reports the hash speed for the CPU mining performance for the selected cryptocurrency. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: MagiEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488140280420560700SE +/- 0.16, N = 3SE +/- 0.22, N = 3SE +/- 1.33, N = 3SE +/- 0.52, N = 3SE +/- 2.02, N = 3SE +/- 0.30, N = 3SE +/- 0.26, N = 3SE +/- 0.39, N = 3453.87617.79248.85469.54328.70330.62633.44278.091. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

Stress-NG

Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Floating PointEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24882K4K6K8K10KSE +/- 1.22, N = 3SE +/- 8.85, N = 3SE +/- 1.77, N = 3SE +/- 2.90, N = 3SE +/- 0.05, N = 3SE +/- 6.00, N = 3SE +/- 15.20, N = 3SE +/- 7.41, N = 35850.597928.203175.545977.764192.474251.398199.654740.281. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -ljpeg -lmpfr -lpthread -lrt -lsctp -lz

Cpuminer-Opt

Cpuminer-Opt is a fork of cpuminer-multi that carries a wide range of CPU performance optimizations for measuring the potential cryptocurrency mining performance of the CPU/processor with a wide variety of cryptocurrencies. The benchmark reports the hash speed for the CPU mining performance for the selected cryptocurrency. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: Myriad-GroestlEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24882K4K6K8K10KSE +/- 39.39, N = 3SE +/- 3.33, N = 3SE +/- 8.61, N = 3SE +/- 5.30, N = 3SE +/- 3.31, N = 3SE +/- 6.11, N = 3SE +/- 13.33, N = 3SE +/- 30.23, N = 38184.4411207.004487.288480.445929.565972.0911337.004843.721. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: DeepcoinEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24882K4K6K8K10KSE +/- 7.14, N = 3SE +/- 18.11, N = 3SE +/- 1.04, N = 3SE +/- 23.29, N = 3SE +/- 1.63, N = 3SE +/- 3.02, N = 3SE +/- 5.77, N = 3SE +/- 36.27, N = 35939.528225.263247.546229.264280.914377.748356.743294.491. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

Stress-NG

Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Vector ShuffleEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24884K8K12K16K20KSE +/- 1.99, N = 3SE +/- 37.39, N = 3SE +/- 3.05, N = 3SE +/- 4.46, N = 3SE +/- 1.05, N = 3SE +/- 2.63, N = 3SE +/- 12.57, N = 3SE +/- 104.90, N = 311892.6916010.746403.8712180.808468.868556.7216452.479138.291. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -ljpeg -lmpfr -lpthread -lrt -lsctp -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Fused Multiply-AddEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248812M24M36M48M60MSE +/- 28005.99, N = 3SE +/- 54521.03, N = 3SE +/- 283045.68, N = 3SE +/- 58656.32, N = 3SE +/- 319716.71, N = 3SE +/- 252666.10, N = 3SE +/- 500794.58, N = 3SE +/- 267691.10, N = 339475809.7855781750.8422079544.9742202851.9729040929.2429927440.6956410032.2529762395.571. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -ljpeg -lmpfr -lpthread -lrt -lsctp -lz

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 16 - Buffer Length: 256 - Filter Length: 32EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488200M400M600M800M1000MSE +/- 956399.50, N = 3SE +/- 1995247.13, N = 3SE +/- 383507.21, N = 3SE +/- 3918917.08, N = 3SE +/- 1997031.13, N = 3SE +/- 812656.82, N = 3SE +/- 67412.49, N = 3SE +/- 5153390.47, N = 37284600007726166675461366677183066677180200007355666677959666677110500001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Stress-NG

Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: AVX-512 VNNIEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488600K1200K1800K2400K3000KSE +/- 97.63, N = 3SE +/- 5242.51, N = 3SE +/- 655.92, N = 3SE +/- 1104.16, N = 3SE +/- 340.15, N = 3SE +/- 1225.91, N = 3SE +/- 2539.18, N = 3SE +/- 5501.16, N = 31758023.972511125.721008567.361909466.061315000.021371242.642571394.821125583.011. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -ljpeg -lmpfr -lpthread -lrt -lsctp -lz

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 1 - Buffer Length: 256 - Filter Length: 32EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248814M28M42M56M70MSE +/- 26514.15, N = 3SE +/- 223381.09, N = 3SE +/- 2905.93, N = 3SE +/- 16522.71, N = 3SE +/- 1855.92, N = 3SE +/- 2645.75, N = 3SE +/- 318365.79, N = 3SE +/- 637003.49, N = 353147000495846675027233351031000522516675365900053018667632366671. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Cpuminer-Opt

Cpuminer-Opt is a fork of cpuminer-multi that carries a wide range of CPU performance optimizations for measuring the potential cryptocurrency mining performance of the CPU/processor with a wide variety of cryptocurrencies. The benchmark reports the hash speed for the CPU mining performance for the selected cryptocurrency. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: GarlicoinEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24887001400210028003500SE +/- 3.95, N = 3SE +/- 0.76, N = 3SE +/- 1.62, N = 3SE +/- 11.09, N = 3SE +/- 2.95, N = 3SE +/- 1.43, N = 3SE +/- 4.61, N = 3SE +/- 4.51, N = 32229.573149.431243.882391.441637.721674.993179.321356.311. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: scryptEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248870140210280350SE +/- 0.11, N = 3SE +/- 1.22, N = 3SE +/- 0.13, N = 3SE +/- 1.04, N = 3SE +/- 0.16, N = 3SE +/- 0.10, N = 3SE +/- 1.34, N = 3SE +/- 0.08, N = 3210.64298.15120.39227.45157.91162.88308.21119.861. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: Quad SHA-256, PyriteEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248813K26K39K52K65KSE +/- 3.33, N = 3SE +/- 454.25, N = 3SE +/- 58.59, N = 3SE +/- 26.67, N = 3SE +/- 38.44, N = 3SE +/- 78.81, N = 3SE +/- 331.71, N = 3SE +/- 3.33, N = 342507600672072045197322933190361360265271. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: RingcoinEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24888001600240032004000SE +/- 2.43, N = 3SE +/- 6.91, N = 3SE +/- 1.54, N = 3SE +/- 2.15, N = 3SE +/- 2.14, N = 3SE +/- 4.79, N = 3SE +/- 20.59, N = 3SE +/- 3.73, N = 32468.433424.571364.322586.651806.871851.253500.131437.471. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: Triple SHA-256, OnecoinEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820K40K60K80K100KSE +/- 72.11, N = 3SE +/- 20.82, N = 3SE +/- 11.55, N = 3SE +/- 188.09, N = 3SE +/- 25.17, N = 3SE +/- 8.82, N = 3SE +/- 146.40, N = 3SE +/- 3.33, N = 367820939703271071517499005108797340388471. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: SkeincoinEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24887K14K21K28K35KSE +/- 42.56, N = 3SE +/- 8.82, N = 3SE +/- 11.55, N = 3SE +/- 134.45, N = 3SE +/- 11.55, N = 3SE +/- 39.30, N = 3SE +/- 11.55, N = 3SE +/- 139.08, N = 324937336931339025583175801798334640139331. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: Blake-2 SEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248830K60K90K120K150KSE +/- 103.98, N = 3SE +/- 41.77, N = 3SE +/- 30.00, N = 3SE +/- 40.41, N = 3SE +/- 20.82, N = 3SE +/- 29.06, N = 3SE +/- 99.39, N = 3SE +/- 460.90, N = 3100853140467538701067907313074513142753561601. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 4 - Input: Bosphorus 4KEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488246810SE +/- 0.019, N = 3SE +/- 0.030, N = 3SE +/- 0.005, N = 3SE +/- 0.004, N = 3SE +/- 0.008, N = 3SE +/- 0.016, N = 3SE +/- 0.009, N = 3SE +/- 0.024, N = 36.5267.2964.5406.4135.5635.7427.8384.3481. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 13 - Input: Bosphorus 4KEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24884080120160200SE +/- 2.97, N = 15SE +/- 4.16, N = 15SE +/- 0.98, N = 15SE +/- 2.98, N = 15SE +/- 1.88, N = 15SE +/- 2.01, N = 15SE +/- 4.27, N = 15SE +/- 0.99, N = 15141.77162.0097.60140.09119.57122.91165.29104.611. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

JPEG-XL libjxl

The JPEG XL Image Coding System is designed to provide next-generation JPEG image capabilities with JPEG XL offering better image quality and compression over legacy JPEG. This test profile is currently focused on the multi-threaded JPEG XL image encode performance using the reference libjxl library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterJPEG-XL libjxl 0.10.1Input: JPEG - Quality: 100EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488612182430SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 320.5325.9112.7019.8116.2816.5227.2213.481. (CXX) g++ options: -fno-rtti -O3 -fPIE -pie -lm

OpenBenchmarking.orgMP/s, More Is BetterJPEG-XL libjxl 0.10.1Input: PNG - Quality: 100EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488714212835SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.07, N = 320.7426.4112.8519.9416.3616.6627.7513.651. (CXX) g++ options: -fno-rtti -O3 -fPIE -pie -lm

Y-Cruncher

Y-Cruncher is a multi-threaded Pi benchmark capable of computing Pi to trillions of digits. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterY-Cruncher 0.8.3Pi Digits To Calculate: 1BEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488816243240SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.07, N = 3SE +/- 0.04, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.01, N = 3SE +/- 0.21, N = 320.2517.7331.3219.3526.1625.4618.2133.89

Timed Mesa Compilation

This test profile times how long it takes to compile Mesa with Meson/Ninja. For minimizing build dependencies and avoid versioning conflicts, test this is just the core Mesa build without LLVM or the extra Gallium3D/Mesa drivers enabled. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Mesa Compilation 24.0Time To CompileEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488816243240SE +/- 0.08, N = 3SE +/- 0.01, N = 3SE +/- 0.16, N = 3SE +/- 0.04, N = 3SE +/- 0.11, N = 3SE +/- 0.11, N = 3SE +/- 0.21, N = 3SE +/- 0.07, N = 322.1217.6735.7121.0528.6228.2617.4632.52

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: H2 Database EngineEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24885001000150020002500SE +/- 23.54, N = 4SE +/- 5.29, N = 3SE +/- 14.53, N = 3SE +/- 9.02, N = 3SE +/- 11.72, N = 3SE +/- 7.51, N = 3SE +/- 13.96, N = 3SE +/- 23.84, N = 319141715239418501992197718262128

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: Very FastEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881020304050SE +/- 0.03, N = 3SE +/- 0.02, N = 4SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.05, N = 3SE +/- 0.08, N = 4SE +/- 0.06, N = 329.8240.1519.2929.5324.6225.2542.6018.97

PyBench

This test profile reports the total time of the different average timed test results from PyBench. PyBench reports average test times for different functions such as BuiltinFunctionCalls and NestedForLoops, with this total result providing a rough estimate as to Python's average performance on a given system. This test profile runs PyBench each time for 20 rounds. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyBench 2018-02-16Total For Average Test TimesEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488130260390520650SE +/- 5.33, N = 4SE +/- 6.56, N = 15SE +/- 2.17, N = 4SE +/- 6.12, N = 4SE +/- 2.21, N = 4SE +/- 2.53, N = 4SE +/- 6.41, N = 4SE +/- 2.82, N = 5526571550602532527533415

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.7Preset: FastEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248890180270360450SE +/- 0.15, N = 3SE +/- 0.31, N = 4SE +/- 0.12, N = 3SE +/- 0.09, N = 3SE +/- 0.39, N = 3SE +/- 0.29, N = 3SE +/- 0.54, N = 4SE +/- 0.04, N = 3258.38365.32147.48259.01192.66203.10392.24205.931. (CXX) g++ options: -O3 -flto -pthread

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: Super FastEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881122334455SE +/- 0.02, N = 3SE +/- 0.05, N = 4SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.07, N = 4SE +/- 0.08, N = 331.6743.8920.3831.2926.0426.6546.4120.17

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: EP.CEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24885001000150020002500SE +/- 21.84, N = 15SE +/- 13.19, N = 8SE +/- 8.87, N = 6SE +/- 11.67, N = 7SE +/- 11.50, N = 10SE +/- 41.08, N = 15SE +/- 13.01, N = 8SE +/- 55.41, N = 151542.112051.601168.351508.631489.701169.682123.371470.811. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 32 - Model: AlexNetEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248850100150200250SE +/- 0.20, N = 3SE +/- 0.35, N = 4SE +/- 0.15, N = 3SE +/- 0.25, N = 3SE +/- 0.06, N = 3SE +/- 0.23, N = 3SE +/- 0.14, N = 4SE +/- 1.37, N = 3204.70239.93139.36199.14171.59172.89242.49160.52

srsRAN Project

srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.10.1-20240325Test: PUSCH Processor Benchmark, Throughput TotalEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248830060090012001500SE +/- 0.14, N = 4SE +/- 18.52, N = 4SE +/- 0.65, N = 6SE +/- 0.05, N = 4SE +/- 0.05, N = 5SE +/- 0.10, N = 6SE +/- 0.05, N = 4SE +/- 9.26, N = 151351.91592.31206.01273.91469.81470.51611.51068.9-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 955 / MAX: 1352.2-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 927.1 / MAX: 1610.9-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 763.8 / MAX: 1207.4-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 849.1 / MAX: 1274-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 849.1 / MAX: 1470-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 849.1 / MAX: 1470.7-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 1019.4 / MAX: 1611.6MIN: 637.2 / MAX: 1111.51. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -fno-trapping-math -fno-math-errno -ldl

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Apache TomcatEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24882K4K6K8K10KSE +/- 4.80, N = 4SE +/- 2.87, N = 4SE +/- 4.16, N = 3SE +/- 5.15, N = 4SE +/- 6.08, N = 3SE +/- 9.21, N = 3SE +/- 2.95, N = 4SE +/- 14.71, N = 344413384862844596485648233706687

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: Deconvolution Batch shapes_1d - Engine: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488246810SE +/- 0.02716, N = 3SE +/- 0.01094, N = 3SE +/- 0.00480, N = 3SE +/- 0.01511, N = 3SE +/- 0.00446, N = 3SE +/- 0.00421, N = 3SE +/- 0.01295, N = 3SE +/- 0.02381, N = 33.575683.113075.214623.345053.934453.910893.006366.85318MIN: 2.86MIN: 2.51MIN: 4.63MIN: 2.98MIN: 3.52MIN: 3.5MIN: 2.4MIN: 3.931. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

7-Zip Compression

This is a test of 7-Zip compression/decompression with its integrated benchmark feature. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 22.01Test: Decompression RatingEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248840K80K120K160K200KSE +/- 184.46, N = 3SE +/- 83.27, N = 3SE +/- 31.18, N = 3SE +/- 503.79, N = 3SE +/- 107.68, N = 3SE +/- 299.70, N = 3SE +/- 369.09, N = 3SE +/- 900.25, N = 3118507167201678681071378840988945170373615271. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 22.01Test: Compression RatingEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248840K80K120K160K200KSE +/- 160.29, N = 3SE +/- 738.67, N = 3SE +/- 155.58, N = 3SE +/- 350.41, N = 3SE +/- 134.28, N = 3SE +/- 195.02, N = 3SE +/- 313.05, N = 3SE +/- 436.66, N = 313993517712582933141565104248103571170888949901. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Ultra FastEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820406080100SE +/- 0.03, N = 5SE +/- 0.04, N = 6SE +/- 0.02, N = 4SE +/- 0.06, N = 5SE +/- 0.04, N = 4SE +/- 0.04, N = 4SE +/- 0.07, N = 6SE +/- 0.30, N = 1057.6874.9641.1156.2850.7152.3081.2339.591. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: Ultra FastEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881224364860SE +/- 0.01, N = 4SE +/- 0.17, N = 4SE +/- 0.01, N = 3SE +/- 0.03, N = 4SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.08, N = 5SE +/- 0.01, N = 337.2251.6324.5936.7131.3031.8654.5424.69

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Apache Lucene Search EngineEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24887001400210028003500SE +/- 18.41, N = 5SE +/- 16.10, N = 5SE +/- 20.21, N = 5SE +/- 7.90, N = 5SE +/- 19.37, N = 5SE +/- 18.64, N = 5SE +/- 9.53, N = 5SE +/- 50.09, N = 1523742127282923132391240920803113

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: IP Shapes 1D - Engine: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24880.65571.31141.96712.62283.2785SE +/- 0.00923, N = 4SE +/- 0.00331, N = 4SE +/- 0.00010, N = 4SE +/- 0.00037, N = 4SE +/- 0.00137, N = 4SE +/- 0.00536, N = 4SE +/- 0.00269, N = 4SE +/- 0.01544, N = 41.973621.166742.606381.935531.979321.999861.130002.91408MIN: 1.82MIN: 1.08MIN: 2.57MIN: 1.91MIN: 1.91MIN: 1.85MIN: 1.03MIN: 2.511. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Super FastEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881428425670SE +/- 0.04, N = 4SE +/- 0.03, N = 5SE +/- 0.04, N = 3SE +/- 0.03, N = 4SE +/- 0.05, N = 4SE +/- 0.05, N = 4SE +/- 0.07, N = 5SE +/- 0.31, N = 444.4658.2530.1043.5437.7038.8262.5628.641. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Very FastEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881122334455SE +/- 0.08, N = 3SE +/- 0.02, N = 4SE +/- 0.02, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.09, N = 4SE +/- 0.19, N = 335.2344.5924.2234.7130.1431.0448.1123.051. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

ASKAP

ASKAP is a set of benchmarks from the Australian SKA Pathfinder. The principal ASKAP benchmarks are the Hogbom Clean Benchmark (tHogbomClean) and Convolutional Resamping Benchmark (tConvolve) as well as some previous ASKAP benchmarks being included as well for OpenCL and CUDA execution of tConvolve. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMillion Grid Points Per Second, More Is BetterASKAP 1.0Test: tConvolve OpenMP - DegriddingEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24882K4K6K8K10KSE +/- 42.40, N = 6SE +/- 90.51, N = 6SE +/- 154.63, N = 7SE +/- 71.64, N = 6SE +/- 18.97, N = 15SE +/- 27.18, N = 8SE +/- 49.85, N = 6SE +/- 41.03, N = 54566.428830.774370.028591.903420.873426.035172.444282.181. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp

OpenBenchmarking.orgMillion Grid Points Per Second, More Is BetterASKAP 1.0Test: tConvolve OpenMP - GriddingEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248814002800420056007000SE +/- 19.10, N = 6SE +/- 55.23, N = 6SE +/- 25.19, N = 7SE +/- 49.58, N = 6SE +/- 21.74, N = 15SE +/- 24.04, N = 8SE +/- 41.40, N = 6SE +/- 25.17, N = 54438.016603.672825.314948.433122.292739.345073.232556.221. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp

LuxCoreRender

LuxCoreRender is an open-source 3D physically based renderer formerly known as LuxRender. LuxCoreRender supports CPU-based rendering as well as GPU acceleration via OpenCL, NVIDIA CUDA, and NVIDIA OptiX interfaces. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: Rainbow Colors and Prism - Acceleration: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248848121620SE +/- 0.05, N = 4SE +/- 0.05, N = 5SE +/- 0.01, N = 3SE +/- 0.03, N = 4SE +/- 0.03, N = 4SE +/- 0.04, N = 4SE +/- 0.06, N = 5SE +/- 0.03, N = 312.7016.438.8812.3011.5911.9316.989.35MIN: 11.78 / MAX: 12.96MIN: 15.06 / MAX: 16.65MIN: 8.44 / MAX: 8.94MIN: 11.45 / MAX: 12.4MIN: 10.79 / MAX: 11.71MIN: 11.1 / MAX: 12.12MIN: 15.46 / MAX: 17.3MIN: 8.52 / MAX: 10.72

SVT-VP9

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-VP9 CPU-based multi-threaded video encoder for the VP9 video format with a sample YUV input video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: VMAF Optimized - Input: Bosphorus 4KEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820406080100SE +/- 0.08, N = 6SE +/- 0.18, N = 6SE +/- 0.11, N = 5SE +/- 0.12, N = 6SE +/- 0.03, N = 5SE +/- 0.09, N = 5SE +/- 0.15, N = 6SE +/- 0.46, N = 1592.07107.5360.9093.5670.3471.48103.5279.851. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

VVenC

VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVVenC 1.11Video Input: Bosphorus 1080p - Video Preset: FasterEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881122334455SE +/- 0.10, N = 4SE +/- 0.12, N = 4SE +/- 0.06, N = 4SE +/- 0.07, N = 4SE +/- 0.12, N = 4SE +/- 0.04, N = 4SE +/- 0.11, N = 4SE +/- 0.25, N = 446.1648.8637.6345.9843.8345.1450.9140.071. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects

Y-Cruncher

Y-Cruncher is a multi-threaded Pi benchmark capable of computing Pi to trillions of digits. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterY-Cruncher 0.8.3Pi Digits To Calculate: 500MEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248848121620SE +/- 0.009, N = 5SE +/- 0.010, N = 5SE +/- 0.082, N = 4SE +/- 0.012, N = 5SE +/- 0.038, N = 4SE +/- 0.036, N = 4SE +/- 0.010, N = 5SE +/- 0.124, N = 49.4148.37314.1338.96911.86111.5898.65214.926

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.7Preset: MediumEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.03, N = 4SE +/- 0.06, N = 5SE +/- 0.02, N = 3SE +/- 0.07, N = 4SE +/- 0.04, N = 4SE +/- 0.04, N = 4SE +/- 0.18, N = 5SE +/- 0.00, N = 3101.80143.3058.08101.6176.0179.60154.7289.141. (CXX) g++ options: -O3 -flto -pthread

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: MediumEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881428425670SE +/- 0.06, N = 4SE +/- 0.10, N = 5SE +/- 0.04, N = 3SE +/- 0.06, N = 4SE +/- 0.03, N = 4SE +/- 0.03, N = 4SE +/- 0.07, N = 5SE +/- 0.14, N = 449.2459.4832.6347.9844.4345.6364.1636.59

Parallel BZIP2 Compression

This test measures the time needed to compress a file (FreeBSD-13.0-RELEASE-amd64-memstick.img) using Parallel BZIP2 compression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterParallel BZIP2 Compression 1.1.13FreeBSD-13.0-RELEASE-amd64-memstick.img CompressionEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488246810SE +/- 0.044427, N = 15SE +/- 0.022219, N = 15SE +/- 0.047798, N = 6SE +/- 0.036366, N = 9SE +/- 0.051557, N = 7SE +/- 0.045809, N = 15SE +/- 0.025811, N = 10SE +/- 0.044479, N = 64.4479043.5215657.6173054.4083235.7018566.1544603.2998796.6124861. (CXX) g++ options: -O2 -pthread -lbz2 -lpthread

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 8 - Input: Bosphorus 4KEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881632486480SE +/- 0.05, N = 4SE +/- 0.11, N = 5SE +/- 0.11, N = 3SE +/- 0.07, N = 4SE +/- 0.07, N = 4SE +/- 0.10, N = 4SE +/- 0.15, N = 5SE +/- 0.30, N = 357.2369.9636.7056.9645.9547.2972.7436.141. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: PMD Source Code AnalyzerEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248830060090012001500SE +/- 7.50, N = 7SE +/- 3.44, N = 7SE +/- 9.62, N = 11SE +/- 9.69, N = 7SE +/- 8.67, N = 11SE +/- 9.55, N = 7SE +/- 5.68, N = 7SE +/- 8.96, N = 1511091051129811011155111310531122

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 1 - Model: ResNet-50EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248848121620SE +/- 0.02, N = 5SE +/- 0.01, N = 5SE +/- 0.02, N = 5SE +/- 0.01, N = 5SE +/- 0.03, N = 5SE +/- 0.01, N = 5SE +/- 0.03, N = 5SE +/- 0.01, N = 514.7615.1215.6114.7617.4517.4714.9313.71

Primesieve

Primesieve generates prime numbers using a highly optimized sieve of Eratosthenes implementation. Primesieve primarily benchmarks the CPU's L1/L2 cache performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterPrimesieve 12.1Length: 1e12EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488510152025SE +/- 0.007, N = 5SE +/- 0.009, N = 6SE +/- 0.014, N = 3SE +/- 0.005, N = 5SE +/- 0.013, N = 4SE +/- 0.007, N = 4SE +/- 0.007, N = 6SE +/- 0.302, N = 39.9437.03016.8279.18312.93712.5066.69121.0951. (CXX) g++ options: -O3

x265

This is a simple test of the x265 encoder run on the CPU with 1080p and 4K options for H.265 video encode performance with x265. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betterx265 3.4Video Input: Bosphorus 1080pEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.15, N = 7SE +/- 0.21, N = 7SE +/- 0.38, N = 6SE +/- 0.18, N = 7SE +/- 0.54, N = 6SE +/- 0.53, N = 6SE +/- 0.21, N = 7SE +/- 0.77, N = 15103.25109.5780.5899.2286.7889.12118.0684.331. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

SVT-HEVC

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 7 - Input: Bosphorus 4KEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820406080100SE +/- 0.05, N = 5SE +/- 0.15, N = 6SE +/- 0.02, N = 4SE +/- 0.07, N = 5SE +/- 0.04, N = 4SE +/- 0.04, N = 5SE +/- 0.06, N = 6SE +/- 0.08, N = 471.6497.7743.5772.0755.6657.2399.6147.981. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt

SVT-VP9

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-VP9 CPU-based multi-threaded video encoder for the VP9 video format with a sample YUV input video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: PSNR/SSIM Optimized - Input: Bosphorus 4KEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 0.11, N = 6SE +/- 0.16, N = 6SE +/- 0.05, N = 5SE +/- 0.15, N = 6SE +/- 0.09, N = 5SE +/- 0.08, N = 5SE +/- 0.13, N = 6SE +/- 0.64, N = 1097.27114.8863.7697.4174.7176.06111.0183.191. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

libavif avifenc

This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 1.0Encoder Speed: 6, LosslessEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24883691215SE +/- 0.018, N = 6SE +/- 0.051, N = 9SE +/- 0.012, N = 5SE +/- 0.038, N = 6SE +/- 0.006, N = 6SE +/- 0.021, N = 6SE +/- 0.044, N = 7SE +/- 0.014, N = 57.2156.17410.0247.1688.2628.1475.8908.9691. (CXX) g++ options: -O3 -fPIC -lm

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 4 - Input: Bosphorus 1080pEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488612182430SE +/- 0.04, N = 6SE +/- 0.08, N = 6SE +/- 0.02, N = 5SE +/- 0.07, N = 6SE +/- 0.04, N = 5SE +/- 0.07, N = 5SE +/- 0.13, N = 6SE +/- 0.15, N = 521.5122.1716.5420.7519.5519.8924.1715.921. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

SVT-HEVC

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 10 - Input: Bosphorus 4KEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24884080120160200SE +/- 0.15, N = 7SE +/- 0.31, N = 8SE +/- 0.10, N = 6SE +/- 0.17, N = 7SE +/- 0.10, N = 6SE +/- 0.12, N = 6SE +/- 0.14, N = 8SE +/- 1.10, N = 15130.38162.4786.79131.05107.47109.03158.94101.831. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: IP Shapes 3D - Engine: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488246810SE +/- 0.00774, N = 5SE +/- 0.01016, N = 5SE +/- 0.00220, N = 5SE +/- 0.00680, N = 5SE +/- 0.00237, N = 5SE +/- 0.00833, N = 5SE +/- 0.00611, N = 5SE +/- 0.01930, N = 53.806932.890874.494062.704304.516294.678403.886857.59695MIN: 3.54MIN: 2.68MIN: 4.45MIN: 2.66MIN: 4.45MIN: 4.46MIN: 3.57MIN: 7.291. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: MediumEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248820406080100SE +/- 0.05, N = 5SE +/- 0.13, N = 6SE +/- 0.05, N = 4SE +/- 0.10, N = 5SE +/- 0.05, N = 5SE +/- 0.04, N = 5SE +/- 0.08, N = 6SE +/- 0.37, N = 466.4780.5247.1763.8562.7164.6888.3050.471. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

Google Draco

Draco is a library developed by Google for compressing/decompressing 3D geometric meshes and point clouds. This test profile uses some Artec3D PLY models as the sample 3D model input formats for Draco compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterGoogle Draco 1.5.6Model: Church FacadeEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248812002400360048006000SE +/- 26.89, N = 7SE +/- 90.20, N = 15SE +/- 16.57, N = 6SE +/- 4.33, N = 6SE +/- 5.95, N = 7SE +/- 10.00, N = 7SE +/- 29.27, N = 6SE +/- 17.70, N = 6503739785160513249885083522456071. (CXX) g++ options: -O3

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: Very FastEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24884080120160200SE +/- 0.05, N = 7SE +/- 0.18, N = 8SE +/- 0.04, N = 6SE +/- 0.08, N = 7SE +/- 0.04, N = 7SE +/- 0.14, N = 7SE +/- 0.11, N = 9SE +/- 0.95, N = 15117.12154.0284.36114.08108.21111.78166.8189.31

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 12 - Input: Bosphorus 4KEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24884080120160200SE +/- 0.26, N = 7SE +/- 0.17, N = 8SE +/- 0.10, N = 6SE +/- 0.18, N = 7SE +/- 0.12, N = 7SE +/- 0.14, N = 7SE +/- 0.33, N = 8SE +/- 1.00, N = 15145.04166.3798.45143.28121.50125.28170.21105.281. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: Convolution Batch Shapes Auto - Engine: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488246810SE +/- 0.01722, N = 7SE +/- 0.00632, N = 7SE +/- 0.03028, N = 7SE +/- 0.00824, N = 7SE +/- 0.01592, N = 7SE +/- 0.00717, N = 7SE +/- 0.01415, N = 7SE +/- 0.04485, N = 76.463574.245327.934554.029367.792028.168006.407196.38053MIN: 6.03MIN: 3.87MIN: 7.8MIN: 3.95MIN: 7.68MIN: 7.7MIN: 5.93MIN: 5.921. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: Super FastEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24884080120160200SE +/- 0.05, N = 8SE +/- 0.07, N = 9SE +/- 0.03, N = 6SE +/- 0.09, N = 8SE +/- 0.05, N = 7SE +/- 0.08, N = 7SE +/- 0.16, N = 9SE +/- 0.97, N = 15127.69168.1388.44124.52116.21119.83181.8396.33

libavif avifenc

This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 1.0Encoder Speed: 6EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488246810SE +/- 0.014, N = 8SE +/- 0.018, N = 9SE +/- 0.010, N = 6SE +/- 0.008, N = 8SE +/- 0.005, N = 7SE +/- 0.014, N = 7SE +/- 0.010, N = 9SE +/- 0.088, N = 154.4373.5686.7984.4695.4365.3383.4216.2601. (CXX) g++ options: -O3 -fPIC -lm

srsRAN Project

srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.10.1-20240325Test: PDSCH Processor Benchmark, Throughput TotalEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24883K6K9K12K15KSE +/- 22.68, N = 6SE +/- 68.24, N = 6SE +/- 13.19, N = 7SE +/- 99.73, N = 6SE +/- 11.93, N = 7SE +/- 38.56, N = 7SE +/- 51.53, N = 6SE +/- 5.11, N = 510640.813807.26218.29792.57583.77530.012852.62908.5-mavx512f -mavx512cd -mavx512bw -mavx512dq-mavx512f -mavx512cd -mavx512bw -mavx512dq-mavx512f -mavx512cd -mavx512bw -mavx512dq-mavx512f -mavx512cd -mavx512bw -mavx512dq-mavx512f -mavx512cd -mavx512bw -mavx512dq-mavx512f -mavx512cd -mavx512bw -mavx512dq-mavx512f -mavx512cd -mavx512bw -mavx512dq1. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -fno-trapping-math -fno-math-errno -ldl

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 8 - Input: Bosphorus 1080pEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24884080120160200SE +/- 0.26, N = 8SE +/- 0.32, N = 8SE +/- 0.16, N = 7SE +/- 0.18, N = 8SE +/- 0.16, N = 8SE +/- 0.13, N = 8SE +/- 0.29, N = 9SE +/- 1.11, N = 15162.42175.21119.76156.68142.84148.34189.41120.531. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

GNU Octave Benchmark

This test profile measures how long it takes to complete several reference GNU Octave files via octave-benchmark. GNU Octave is used for numerical computations and is an open-source alternative to MATLAB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterGNU Octave Benchmark 8.4.0EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488246810SE +/- 0.018, N = 7SE +/- 0.020, N = 7SE +/- 0.013, N = 7SE +/- 0.016, N = 7SE +/- 0.016, N = 7SE +/- 0.017, N = 7SE +/- 0.021, N = 7SE +/- 0.011, N = 65.8155.4805.6945.5265.5835.5875.7806.462

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: Ultra FastEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24884080120160200SE +/- 0.12, N = 8SE +/- 0.12, N = 9SE +/- 0.06, N = 7SE +/- 0.14, N = 8SE +/- 0.16, N = 8SE +/- 0.06, N = 8SE +/- 0.20, N = 9SE +/- 1.21, N = 15150.51187.73105.23146.08138.95142.97203.39115.40

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: Very FastEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24884080120160200SE +/- 0.06, N = 8SE +/- 0.11, N = 9SE +/- 0.06, N = 7SE +/- 0.16, N = 8SE +/- 0.09, N = 7SE +/- 0.10, N = 8SE +/- 0.25, N = 9SE +/- 0.87, N = 7131.13169.81101.80125.34123.89129.02186.56103.041. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Zxing 1D/2D Barcode Image ProcessingEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248830060090012001500SE +/- 2.11, N = 9SE +/- 3.84, N = 9SE +/- 10.99, N = 7SE +/- 3.68, N = 9SE +/- 2.15, N = 8SE +/- 2.37, N = 8SE +/- 4.09, N = 9SE +/- 30.56, N = 157816201380749104110195891053

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: Super FastEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248850100150200250SE +/- 0.12, N = 9SE +/- 0.14, N = 10SE +/- 0.06, N = 8SE +/- 0.10, N = 9SE +/- 0.04, N = 9SE +/- 0.09, N = 9SE +/- 0.26, N = 10SE +/- 2.29, N = 15172.91224.02126.08166.31161.51168.08248.39139.721. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

SVT-HEVC

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 7 - Input: Bosphorus 1080pEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248870140210280350SE +/- 0.24, N = 9SE +/- 0.40, N = 10SE +/- 0.13, N = 7SE +/- 0.22, N = 9SE +/- 0.17, N = 8SE +/- 0.21, N = 8SE +/- 0.26, N = 10SE +/- 2.05, N = 15216.67297.03135.99215.68173.19178.20311.95168.391. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt

Helsing

Helsing is an open-source POSIX vampire number generator. This test profile measures the time it takes to generate vampire numbers between varying numbers of digits. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterHelsing 1.0-betaDigit Range: 12 digitEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881.16512.33023.49534.66045.8255SE +/- 0.003, N = 9SE +/- 0.003, N = 11SE +/- 0.002, N = 7SE +/- 0.003, N = 10SE +/- 0.003, N = 8SE +/- 0.004, N = 8SE +/- 0.003, N = 11SE +/- 0.028, N = 153.0042.1175.1782.7713.9663.8282.0164.0801. (CC) gcc options: -O2 -pthread

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 1 - Model: GoogLeNetEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881428425670SE +/- 0.08, N = 8SE +/- 0.18, N = 8SE +/- 0.15, N = 8SE +/- 0.18, N = 8SE +/- 0.09, N = 8SE +/- 0.10, N = 9SE +/- 0.18, N = 8SE +/- 0.05, N = 854.4960.1452.3655.8961.8862.9957.3451.33

LAMMPS Molecular Dynamics Simulator

LAMMPS is a classical molecular dynamics code, and an acronym for Large-scale Atomic/Molecular Massively Parallel Simulator. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 23Jun2022Model: Rhodopsin ProteinEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24883691215SE +/- 0.178, N = 15SE +/- 0.045, N = 11SE +/- 0.004, N = 9SE +/- 0.126, N = 15SE +/- 0.035, N = 10SE +/- 0.055, N = 9SE +/- 0.059, N = 11SE +/- 0.250, N = 159.26611.7666.8389.5258.7156.82612.4817.2361. (CXX) g++ options: -O3 -lm -ldl

SVT-VP9

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-VP9 CPU-based multi-threaded video encoder for the VP9 video format with a sample YUV input video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: Visual Quality Optimized - Input: Bosphorus 1080pEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248880160240320400SE +/- 1.08, N = 10SE +/- 2.24, N = 11SE +/- 0.64, N = 9SE +/- 1.06, N = 10SE +/- 0.69, N = 9SE +/- 0.76, N = 9SE +/- 1.71, N = 15SE +/- 3.49, N = 15284.34357.60173.33281.56198.42203.16364.90221.331. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: Deconvolution Batch shapes_3d - Engine: CPUEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881.24932.49863.74794.99726.2465SE +/- 0.00652, N = 9SE +/- 0.00476, N = 9SE +/- 0.00099, N = 9SE +/- 0.00129, N = 9SE +/- 0.00062, N = 9SE +/- 0.00661, N = 9SE +/- 0.00405, N = 9SE +/- 0.10597, N = 153.454612.707825.552543.510653.948883.834122.461594.94792MIN: 3.24MIN: 2.61MIN: 5.53MIN: 3.46MIN: 3.84MIN: 3.58MIN: 2.35MIN: 4.451. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

srsRAN Project

srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.10.1-20240325Test: PDSCH Processor Benchmark, Throughput ThreadEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24882004006008001000SE +/- 7.68, N = 9SE +/- 3.58, N = 9SE +/- 6.27, N = 10SE +/- 1.53, N = 9SE +/- 7.98, N = 10SE +/- 3.24, N = 10SE +/- 5.92, N = 10SE +/- 5.58, N = 91034.9907.51073.6908.81071.01061.51049.1809.6-mavx512f -mavx512cd -mavx512bw -mavx512dq-mavx512f -mavx512cd -mavx512bw -mavx512dq-mavx512f -mavx512cd -mavx512bw -mavx512dq-mavx512f -mavx512cd -mavx512bw -mavx512dq-mavx512f -mavx512cd -mavx512bw -mavx512dq-mavx512f -mavx512cd -mavx512bw -mavx512dq-mavx512f -mavx512cd -mavx512bw -mavx512dq1. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -fno-trapping-math -fno-math-errno -ldl

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: Ultra FastEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248870140210280350SE +/- 0.14, N = 10SE +/- 0.34, N = 10SE +/- 0.11, N = 9SE +/- 0.16, N = 10SE +/- 0.14, N = 10SE +/- 0.14, N = 10SE +/- 0.21, N = 11SE +/- 4.30, N = 15231.05271.84167.39220.31217.52226.13301.57197.511. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

SVT-VP9

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-VP9 CPU-based multi-threaded video encoder for the VP9 video format with a sample YUV input video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: VMAF Optimized - Input: Bosphorus 1080pEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248890180270360450SE +/- 0.25, N = 10SE +/- 1.10, N = 11SE +/- 0.18, N = 9SE +/- 0.34, N = 10SE +/- 0.18, N = 10SE +/- 0.20, N = 10SE +/- 0.58, N = 11SE +/- 0.18, N = 10331.83410.42208.12330.16237.22242.58408.42289.261. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: PSNR/SSIM Optimized - Input: Bosphorus 1080pEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248890180270360450SE +/- 0.37, N = 11SE +/- 0.58, N = 11SE +/- 0.22, N = 9SE +/- 0.48, N = 10SE +/- 0.35, N = 10SE +/- 0.28, N = 10SE +/- 0.42, N = 11SE +/- 0.53, N = 10337.46419.93213.44336.78246.28251.44419.26295.691. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

SVT-HEVC

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 10 - Input: Bosphorus 1080pEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488120240360480600SE +/- 0.29, N = 11SE +/- 0.81, N = 12SE +/- 0.23, N = 10SE +/- 0.39, N = 11SE +/- 0.44, N = 11SE +/- 0.24, N = 11SE +/- 1.04, N = 12SE +/- 4.53, N = 15428.68554.16279.57422.86351.91361.05560.81383.641. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 12 - Input: Bosphorus 1080pEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488140280420560700SE +/- 1.70, N = 12SE +/- 1.39, N = 12SE +/- 0.74, N = 12SE +/- 0.82, N = 12SE +/- 0.93, N = 12SE +/- 0.74, N = 12SE +/- 1.66, N = 15SE +/- 0.76, N = 12581.67631.94449.57568.41528.23546.71660.34548.041. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

Timed CPython Compilation

This test times how long it takes to build the reference Python implementation, CPython, with optimizations and LTO enabled for a release build. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed CPython Compilation 3.10.6Build Configuration: DefaultEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248851015202515.0614.5918.8515.2116.8216.5814.1316.85

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.4Test: r2c - Backend: FFTW - Precision: double - X Y Z: 128EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24881122334455SE +/- 0.59, N = 15SE +/- 1.39, N = 15SE +/- 0.06, N = 14SE +/- 0.52, N = 15SE +/- 0.13, N = 14SE +/- 0.15, N = 15SE +/- 0.58, N = 15SE +/- 0.39, N = 1535.0250.2827.8322.7024.3823.7237.2426.481. (CXX) g++ options: -O3

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.4Test: r2c - Backend: FFTW - Precision: float - X Y Z: 128EPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-2488306090120150SE +/- 1.06, N = 15SE +/- 1.72, N = 15SE +/- 2.17, N = 15SE +/- 0.82, N = 15SE +/- 1.45, N = 15SE +/- 1.52, N = 15SE +/- 1.14, N = 15SE +/- 3.73, N = 15107.74139.0676.44113.7476.2877.52127.9385.011. (CXX) g++ options: -O3

System Power Consumption Monitor

OpenBenchmarking.orgWattsSystem Power Consumption MonitorPhoronix Test Suite System MonitoringEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248860120180240300Min: 84.9 / Avg: 130.63 / Max: 143.7Min: 76.2 / Avg: 149.2 / Max: 222.4Min: 51.6 / Avg: 115.02 / Max: 143.1Min: 53.5 / Avg: 128.7 / Max: 186.1Min: 51.9 / Avg: 120.42 / Max: 144Min: 67.4 / Avg: 159.31 / Max: 212.3Min: 90.7 / Avg: 202.74 / Max: 301Min: 51.7 / Avg: 138.43 / Max: 338.5

CPU Power Consumption Monitor

OpenBenchmarking.orgWattsCPU Power Consumption MonitorPhoronix Test Suite System MonitoringEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-24884080120160200Min: 46.51 / Avg: 84.39 / Max: 91.68Min: 37.24 / Avg: 95.52 / Max: 147.11Min: 12.24 / Avg: 71.17 / Max: 91.98Min: 16.44 / Avg: 80.5 / Max: 123.57Min: 13.54 / Avg: 75.17 / Max: 92.35Min: 29.45 / Avg: 105.15 / Max: 145.99Min: 48.93 / Avg: 138.83 / Max: 208.13Min: 18.57 / Avg: 89.22 / Max: 198.5

CPU Peak Freq (Highest CPU Core Frequency) Monitor

OpenBenchmarking.orgMegahertzCPU Peak Freq (Highest CPU Core Frequency) MonitorPhoronix Test Suite System MonitoringEPYC 4464PEPYC 4584PXEPYC 4244PEPYC 4484PXEPYC 4344PEPYC 4364PEPYC 4564PXeon E-248810002000300040005000Min: 3947 / Avg: 4915.82 / Max: 5476Min: 4296 / Avg: 4927.95 / Max: 5702Min: 400 / Avg: 4805.55 / Max: 5230Min: 400 / Avg: 4831.8 / Max: 5638Min: 400 / Avg: 4794.94 / Max: 5417Min: 4543 / Avg: 5178.15 / Max: 5518Min: 4577 / Avg: 5148.01 / Max: 5604Min: 801 / Avg: 4466.33 / Max: 5638

367 Results Shown

LAMMPS Molecular Dynamics Simulator
TensorFlow
OpenRadioss
Timed Linux Kernel Compilation
OpenRadioss
OpenFOAM:
  drivaerFastback, Medium Mesh Size - Execution Time
  drivaerFastback, Medium Mesh Size - Mesh Time
TensorFlow
Blender
Timed Node.js Compilation
OpenVKL
TensorFlow
Timed LLVM Compilation
OpenRadioss
Timed LLVM Compilation
Timed Gem5 Compilation
OpenRadioss:
  Rubber O-Ring Seal Installation
  Bumper Beam
MariaDB
Helsing
Apache IoTDB:
  800 - 100 - 800 - 400:
    Average Latency
    point/sec
GPAW
OpenSSL:
  RSA4096:
    verify/s
    sign/s
Apache IoTDB:
  800 - 100 - 800 - 100:
    Average Latency
    point/sec
OpenRadioss
TensorFlow
Blender
Timed Godot Game Engine Compilation
MariaDB:
  oltp_read_write - 64
  oltp_write_only - 64
  oltp_update_index - 64
Stockfish
ClickHouse:
  100M Rows Hits Dataset, Third Run
  100M Rows Hits Dataset, Second Run
  100M Rows Hits Dataset, First Run / Cold Cache
Blender
TensorFlow:
  CPU - 64 - ResNet-50
  CPU - 512 - AlexNet
ONNX Runtime:
  ResNet50 v1-12-int8 - CPU - Standard:
    Inference Time Cost (ms)
    Inferences Per Second
miniBUDE:
  OpenMP - BM2:
    Billion Interactions/s
    GFInst/s
Apache IoTDB:
  800 - 100 - 500 - 400:
    Average Latency
    point/sec
  500 - 100 - 800 - 400:
    Average Latency
    point/sec
  800 - 100 - 500 - 100:
    Average Latency
    point/sec
BRL-CAD
Apache IoTDB:
  500 - 100 - 800 - 100:
    Average Latency
    point/sec
SVT-HEVC
OpenSSL:
  ChaCha20
  AES-256-GCM
  AES-128-GCM
  ChaCha20-Poly1305
  SHA512
  SHA256
MariaDB mariadb-slap
Speedb
Xcompact3d Incompact3d
PostgreSQL:
  100 - 800 - Read Only - Average Latency
  100 - 800 - Read Only
OSPRay
Java JMH
Zstd Compression
PostgreSQL:
  100 - 800 - Read Write - Average Latency
  100 - 800 - Read Write
Appleseed
ScyllaDB
ONNX Runtime:
  fcn-resnet101-11 - CPU - Parallel:
    Inference Time Cost (ms)
    Inferences Per Second
ASTC Encoder:
  Very Thorough
  Exhaustive
Apache IoTDB:
  500 - 100 - 500 - 400:
    Average Latency
    point/sec
OSPRay
Primesieve
Apache IoTDB:
  500 - 100 - 500 - 100:
    Average Latency
    point/sec
PostgreSQL:
  100 - 500 - Read Only - Average Latency
  100 - 500 - Read Only
  100 - 500 - Read Write - Average Latency
  100 - 500 - Read Write
ONNX Runtime:
  yolov4 - CPU - Parallel:
    Inference Time Cost (ms)
    Inferences Per Second
NAS Parallel Benchmarks
FFmpeg:
  libx265 - Platform
  libx265 - Video On Demand
  libx265 - Upload
NAS Parallel Benchmarks
DuckDB
ONNX Runtime:
  T5 Encoder - CPU - Standard:
    Inference Time Cost (ms)
    Inferences Per Second
LuxCoreRender
Blender
SPECFEM3D:
  Water-layered Halfspace
  Layered Halfspace
OSPRay
ONNX Runtime:
  GPT-2 - CPU - Standard:
    Inference Time Cost (ms)
    Inferences Per Second
Intel Open Image Denoise
Blender
TensorFlow
NAMD
TensorFlow
libavif avifenc
CockroachDB:
  KV, 50% Reads - 128
  KV, 60% Reads - 128
  KV, 95% Reads - 128
ONNX Runtime:
  bertsquad-12 - CPU - Parallel:
    Inference Time Cost (ms)
    Inferences Per Second
VVenC
GROMACS
NAS Parallel Benchmarks
Numpy Benchmark
OSPRay Studio
ACES DGEMM
PyTorch
Zstd Compression
Appleseed:
  Material Tester
  Disney Material
ASKAP:
  tConvolve MPI - Gridding
  tConvolve MPI - Degridding
MariaDB mariadb-slap
Blender
nginx:
  1000
  500
  200
OpenFOAM:
  drivaerFastback, Small Mesh Size - Execution Time
  drivaerFastback, Small Mesh Size - Mesh Time
Timed Linux Kernel Compilation
Pennant
OSPRay Studio
oneDNN
CloverLeaf
Zstd Compression
ONNX Runtime:
  CaffeNet 12-int8 - CPU - Parallel:
    Inference Time Cost (ms)
    Inferences Per Second
Zstd Compression
oneDNN
Liquid-DSP
NAMD
OSPRay Studio
TensorFlow
SPECFEM3D
Chaos Group V-RAY
OSPRay Studio
ONNX Runtime:
  ArcFace ResNet-100 - CPU - Parallel:
    Inference Time Cost (ms)
    Inferences Per Second
OSPRay Studio:
  1 - 1080p - 32 - Path Tracer - CPU
  1 - 1080p - 1 - Path Tracer - CPU
  2 - 1080p - 1 - Path Tracer - CPU
Pennant
OSPRay
Timed CPython Compilation
Memcached:
  1:10
  1:100
uvg266
OSPRay:
  gravity_spheres_volume/dim_512/ao/real_time
  gravity_spheres_volume/dim_512/scivis/real_time
ONNX Runtime:
  Faster R-CNN R-50-FPN-int8 - CPU - Parallel:
    Inference Time Cost (ms)
    Inferences Per Second
ASTC Encoder
LuxCoreRender
ONNX Runtime:
  ResNet50 v1-12-int8 - CPU - Parallel:
    Inference Time Cost (ms)
    Inferences Per Second
SPECFEM3D
OSPRay Studio
Zstd Compression
OpenVINO:
  Face Detection FP16 - CPU:
    ms
    FPS
LuxCoreRender
OpenVINO:
  Face Detection FP16-INT8 - CPU:
    ms
    FPS
OSPRay Studio
ONNX Runtime:
  GPT-2 - CPU - Parallel:
    Inference Time Cost (ms)
    Inferences Per Second
GraphicsMagick
LuxCoreRender
OpenVINO:
  Machine Translation EN To DE FP16 - CPU:
    ms
    FPS
ONNX Runtime:
  T5 Encoder - CPU - Parallel:
    Inference Time Cost (ms)
    Inferences Per Second
OpenVINO:
  Person Detection FP16 - CPU:
    ms
    FPS
  Person Detection FP32 - CPU:
    ms
    FPS
GraphicsMagick
OpenVINO:
  Noise Suppression Poconet-Like FP16 - CPU:
    ms
    FPS
  Road Segmentation ADAS FP16-INT8 - CPU:
    ms
    FPS
  Person Vehicle Bike Detection FP16 - CPU:
    ms
    FPS
GraphicsMagick
OpenVINO:
  Road Segmentation ADAS FP16 - CPU:
    ms
    FPS
  Person Re-Identification Retail FP16 - CPU:
    ms
    FPS
  Handwritten English Recognition FP16-INT8 - CPU:
    ms
    FPS
  Vehicle Detection FP16-INT8 - CPU:
    ms
    FPS
  Handwritten English Recognition FP16 - CPU:
    ms
    FPS
  Face Detection Retail FP16-INT8 - CPU:
    ms
    FPS
SPECFEM3D
OpenVINO:
  Age Gender Recognition Retail 0013 FP16-INT8 - CPU:
    ms
    FPS
  Vehicle Detection FP16 - CPU:
    ms
    FPS
  Weld Porosity Detection FP16 - CPU:
    ms
    FPS
GraphicsMagick
OpenVINO:
  Age Gender Recognition Retail 0013 FP16 - CPU:
    ms
    FPS
  Weld Porosity Detection FP16-INT8 - CPU:
    ms
    FPS
  Face Detection Retail FP16 - CPU:
    ms
    FPS
John The Ripper
Speedb
GraphicsMagick:
  HWB Color Space
  Resizing
Speedb
RocksDB:
  Read Rand Write Rand
  Read While Writing
ONNX Runtime:
  super-resolution-10 - CPU - Parallel:
    Inference Time Cost (ms)
    Inferences Per Second
John The Ripper
RocksDB
libavif avifenc
Xcompact3d Incompact3d
Intel Open Image Denoise:
  RT.hdr_alb_nrm.3840x2160 - CPU-Only
  RT.ldr_alb_nrm.3840x2160 - CPU-Only
Z3 Theorem Prover
Timed PHP Compilation
PyTorch
m-queens
Embree
Coremark
Renaissance
Kvazaar
SVT-HEVC
VVenC
x265
FFmpeg
JPEG-XL libjxl:
  JPEG - 90
  PNG - 90
Embree
QuantLib
Timed FFmpeg Compilation
TensorFlow
NAS Parallel Benchmarks
Embree
Timed Wasmer Compilation
SVT-VP9
srsRAN Project
TensorFlow
miniBUDE:
  OpenMP - BM1:
    Billion Interactions/s
    GFInst/s
VVenC
Aircrack-ng
John The Ripper:
  WPA PSK
  bcrypt
  Blowfish
Liquid-DSP
Cpuminer-Opt
Stress-NG:
  Wide Vector Math
  Vector Floating Point
Cpuminer-Opt
Stress-NG
Cpuminer-Opt:
  Myriad-Groestl
  Deepcoin
Stress-NG:
  Vector Shuffle
  Fused Multiply-Add
Liquid-DSP
Stress-NG
Liquid-DSP
Cpuminer-Opt:
  Garlicoin
  scrypt
  Quad SHA-256, Pyrite
  Ringcoin
  Triple SHA-256, Onecoin
  Skeincoin
  Blake-2 S
SVT-AV1:
  Preset 4 - Bosphorus 4K
  Preset 13 - Bosphorus 4K
JPEG-XL libjxl:
  JPEG - 100
  PNG - 100
Y-Cruncher
Timed Mesa Compilation
DaCapo Benchmark
uvg266
PyBench
ASTC Encoder
uvg266
NAS Parallel Benchmarks
TensorFlow
srsRAN Project
DaCapo Benchmark
oneDNN
7-Zip Compression:
  Decompression Rating
  Compression Rating
Kvazaar
uvg266
DaCapo Benchmark
oneDNN
Kvazaar:
  Bosphorus 4K - Super Fast
  Bosphorus 4K - Very Fast
ASKAP:
  tConvolve OpenMP - Degridding
  tConvolve OpenMP - Gridding
LuxCoreRender
SVT-VP9
VVenC
Y-Cruncher
ASTC Encoder
uvg266
Parallel BZIP2 Compression
SVT-AV1
DaCapo Benchmark
TensorFlow
Primesieve
x265
SVT-HEVC
SVT-VP9
libavif avifenc
SVT-AV1
SVT-HEVC
oneDNN
Kvazaar
Google Draco
uvg266
SVT-AV1
oneDNN
uvg266
libavif avifenc
srsRAN Project
SVT-AV1
GNU Octave Benchmark
uvg266
Kvazaar
DaCapo Benchmark
Kvazaar
SVT-HEVC
Helsing
TensorFlow
LAMMPS Molecular Dynamics Simulator
SVT-VP9
oneDNN
srsRAN Project
Kvazaar
SVT-VP9:
  VMAF Optimized - Bosphorus 1080p
  PSNR/SSIM Optimized - Bosphorus 1080p
SVT-HEVC
SVT-AV1
Timed CPython Compilation
HeFFTe - Highly Efficient FFT for Exascale:
  r2c - FFTW - double - 128
  r2c - FFTW - float - 128
System Power Consumption Monitor:
  Phoronix Test Suite System Monitoring:
    Watts
    Watts
    Megahertz